{"id":"5d998796-82ee-4335-932f-70c5810bb0c4","arxiv_id":"2504.20247","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"RT-TDDFTB is presented as a tractable quantum method for studying plasmons in metal nanoparticles up to thousands of atoms, with honest documentation of its failures for alkali metals, nanogap tunneling, and long-time relaxation.","lead":"This paper argues that density functional tight-binding (DFTB), especially its real-time time-dependent form (RT-TDDFTB), can simulate quantum plasmonic effects in nanoparticles that are too large for full time-dependent density functional theory. It is a feature article that reviews and demonstrates the method's use for optical absorption, harmonic generation, and plasmon-driven H2 dissociation.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central 'reasonable accuracy / picosecond' claim is not quantitatively supported: the paper's own Na results show a wrong size trend, and the photocatalysis section limits meaningful dynamics to tens of femtoseconds.","rationale":"The reader flagged the DFTB parameterization and excited-state assumption as the weakest point; I agree, and the paper's own Fig. 2 and its photocatalysis limitation passages are direct evidence of the same soft spot. The paper is honest about limitations and useful as a review, so no rejection is warranted. However, the abstract and conclusion state a broad capability that the evidence does not yet support: no quantitative error statistics are reported, the one direct comparison with a classical model (Fig. 4) is rescaled to force agreement, and the flagship H2 dissociation simulation is limited to 100 fs with a three-orders-of-magnitude discrepancy from experimental thresholds that the authors attribute to missing relaxation. My verdict does not change the reader's CONDITIONAL recommendation, but it sharpens the condition: the article should either add a benchmark table against higher-level theory or experiment, or explicitly restrict the central claim to linear optical properties of noble-metal clusters on sub-100-fs timescales.","tokens_in":37504,"tokens_out":5265,"duration_ms":59433,"concrete_test":"Build a benchmark set of icosahedral or tetrahedral Ag_n, Au_n, and Na_n clusters (n approximately 19, 55, 147, 309, 489) and run RT-TDDFTB with the exact Slater-Koster sets used in the paper (hyb-0-2 for Ag, auorg-1-1 for Au, matsci-0-3 for Na). Compare the dominant absorption peak energy and its size slope against published TDDFT and experimental spectra (e.g., Refs. 102, 126–128). If the mean absolute peak error exceeds about 0.3 eV for Ag/Au, or if the sign of the size trend is wrong for any free-electron metal, then 'reasonable accuracy' must be explicitly scoped to noble metals and specific sizes rather than stated as a general capability.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The claim that RT-TDDFTB enables tractable and reasonably accurate studies of quantum plasmonics rests on two conditions: (i) the tight-binding parameterization preserves the relevant excited-state response, and (ii) the real-time Ehrenfest propagation captures the dynamics that are actually claimed. Tractability is demonstrated, but accuracy is not benchmarked, and the paper itself provides counterexamples to both conditions. In the section 'DFTB Fails to Account for s-Electron Spillover', DFTB-calculated Na13–Na561 spectra absorb in the deep UV and redshift with size, whereas the nonlocal hydrodynamic model (Eqs. 18–21) and higher-level KS-DFT predict visible absorption with a blueshift; the text attributes this to artificial orbital compression and the minimal basis set. A size-dependent plasmon shift is one of the headline capabilities, so this is a qualitative failure, not a small error. For dynamics, the photocatalysis section states that 'the use of RT-TDDFTB outright is limited to the first tens of femtoseconds' because Ehrenfest dynamics does not fully describe electron-electron and electron-phonon scattering; the Al nanoparticle example shows step-like non-thermal distributions persisting rather than relaxing. The abstract's 'picosecond timescales' is therefore a statement about propagation cost, not about validated excited-state dynamics. Absent a quantitative benchmark against TDDFT or experiment for the noble-metal systems emphasized, 'reasonable accuracy' remains an assertion, and the manuscript's own scope limitations should be carried into the central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper argues that real-time time-dependent density functional tight-binding (RT-TDDFTB) provides a tractable quantum-mechanical approach to plasmonic nanostructures with hundreds to thousands of atoms, over timescales claimed to reach picoseconds. It reviews the DFTB formalism (SCC-DFTB2/3, Slater-Koster parameterization), the real-time propagation scheme (Ehrenfest dynamics, Liouville-von Neumann equation), and applications to linear absorption of metal clusters, size-dependent plasmon shifts, classical-quantum reconciliation, plasmon coupling in nanogaps, nonlinear harmonic generation, and plasmon-driven H2 dissociation and hot-carrier dynamics. The paper includes new illustrative DFTB calculations (Na clusters, Ag clusters, Ag nanocube arrays, Na nanocube) and discusses limitations, notably failure to capture s-electron spillover in alkali metals, underestimation of charge-transfer tunneling, and the restriction of Ehrenfest dynamics to the first tens of femtoseconds for photochemical dynamics.","tokens_in":37710,"tokens_out":7580,"duration_ms":71040,"significance":"If the paper's claims were quantitatively supported, the review would be significant because it consolidates a growing literature and points to a practical tool for quantum plasmonics beyond TDDFT's system-size limits. The authors are transparent about several failure modes and suggest concrete methodological improvements (range-separated hybrids, GW, Boltzmann transport). However, the current evidence base is largely qualitative or drawn from the authors' prior work, and the abstract overstates the method's validated accuracy and accessible timescales. With appropriate tempering and explicit benchmarking, the paper could serve as a useful roadmap for the community.","major_comments":[{"comment":"The abstract claims that RT-TDDFTB enables modeling 'over picosecond timescales' and 'capturing key phenomena such as size-dependent plasmon shifts,' but the Photocatalysis section explicitly states that 'the use of RT-TDDFTB outright is limited to the first tens of femtoseconds when analyzing the electron dynamics associated with plasmon-driven photochemistry,' and the Na-cluster results (Fig. 2) show that DFTB predicts the wrong sign of the size-dependent shift for alkali metals. These statements need to be reconciled so the abstract accurately represents the validated scope of the method.","section":"Abstract; Photocatalysis section"},{"comment":"The Na1241 cube comparison is weakened by the frequency scaling described in the Fig. 4 caption: the DFTB spectrum was scaled so that its first peak coincides with the classical reference. As presented, this does not validate the absolute resonance positions and effectively fits the first peak; please show the unscaled spectra or explicitly label the comparison as shape-only.","section":"Fig. 4 (Classical-quantum comparison)"},{"comment":"The claim that DFTB captures d-orbital screening 'with reasonable accuracy' is supported by a qualitative trend (blueshift with decreasing size) but no quantitative comparison to TDDFT or experimental spectra for the same clusters is provided. A benchmark of peak positions or cross sections against higher-level theory or experiment is needed to support the 'reasonable accuracy' wording, or the claim should be softened to 'qualitative agreement.'","section":"Fig. 3 (Ag spectra)"},{"comment":"The statement that DFTB 'may ultimately reveal more about plasmonic phenomena than their more sophisticated counterparts' goes beyond the evidence presented, given the acknowledged failures for alkali-metal spillout, charge-transfer tunneling, and long-time electronic relaxation. This is acceptable as a perspective, but it should be explicitly labeled as an opinion rather than a conclusion supported by the results in this manuscript.","section":"Conclusions; 'High Computational Tractability' section"}],"minor_comments":[{"comment":"There are typos, e.g., 'phemonena' (Concluding section) and 'resonable' (Hierarchical Nanostructures section, discussion of hyperpolarizabilities); these should be corrected.","section":"Throughout"},{"comment":"The phrase 'scalable, parameter-free studies' is misleading because the preceding section emphasizes that DFTB is a 'highly parameterized semiempirical method'; consider replacing 'parameter-free' with 'parameterized' or 'without additional free parameters beyond the Slater-Koster files.'","section":"Hierarchical Nanostructures section"},{"comment":"The label 'Ag666' is ambiguous; please clarify whether this is a 666-atom cluster or a typo.","section":"Fig. 5"},{"comment":"Refs. 113 and 193 are cited as arXiv preprints; please indicate their publication status in the reference list.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is effectively a review of the authors' own prior work (Refs. 85, 113, 181) with some new illustrative data. For a Feature Article this is acceptable, but the editor may want to confirm that the new contributions (Na and Ag spectra, nanocube coupling) are clearly delineated and that the self-citation density in the applications sections is appropriate. No evidence of misconduct; the limitations are openly discussed, which is a strength."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a Feature Article from Schatz's group on RT-TDDFTB for quantum plasmonics. Read it as a review with illustrative simulations, not a new result. The paper does a genuinely useful job: it lays out the DFTB theory (third-order expansion, Slater-Koster integrals, Ehrenfest propagation) and, more valuably, it documents the method's failure modes clearly. The sodium section is the standout: DFTB predicts absorption in the deep UV and a redshift with particle size, while the nonlocal hydrodynamic model and KS-DFT both give visible absorption with a blueshift; the text attributes this to artificial orbital compression and the minimal basis set. The paper also admits that long-range excited-state charge transport through nanogaps is 'incapable' of being fully described, and that RT-TDDFTB electron dynamics are 'limited to the first tens of femtoseconds' for photochemistry. That honesty earns the paper a serious read.\n\nThe soft spots are mostly about the frame. The abstract promises 'thousands of atoms over picosecond timescales' and 'reasonable accuracy.' The picosecond claim is about propagation cost, not validated electron dynamics; the paper's own results show step-like non-thermal electron distributions persisting rather than relaxing. 'Reasonable accuracy' is never benchmarked quantitatively against TDDFT or experiment for the noble-metal cases. The one quantitative-looking comparison, the Na1241 cube against a classical cube spectrum, uses a frequency-scaling factor to align the first peak; that weakens the comparison considerably. Damping and broadening times are also chosen, not derived. So the headline capability is asserted rather than demonstrated.\n\nCitation pattern: as expected for a feature article, heavily self-referential; most illustrative spectra come from the group's own prior papers (Refs. 85, 113, 181). That doesn't make the review circular, but it does mean the endorsement is not independent.\n\nFor a reader new to quantum plasmonics, this is a good map of where DFTB works (Ag, Au, short times, linear and nonlinear response) and where it fails (alkali spillout, tunneling gaps, long-time relaxation). I'd send it to a referee if it were submitted as a review, with the condition that the abstract be rewritten to match the body. The central program is defensible; the overclaim is not.\n\nRecommendation: engage with the paper, cite it as a review of limitations, and treat 'reasonable accuracy / picosecond' as an open question, not a result.","headline":"An honest, readable review of RT-TDDFTB for quantum plasmonics that clearly documents the method's real failures, but the abstract oversells accuracy and timescales; useful as a map, not as a benchmark.","tokens_in":38361,"tokens_out":3668,"would_cite":false,"duration_ms":38976,"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":"Tight-binding method scales quantum plasmonics to thousands of atoms","keywords":["quantum plasmonics","density functional tight-binding","real-time time-dependent DFTB","hot carriers","plasmonic photocatalysis","harmonic generation","plasmon coupling","nanoparticle dimers"],"falsifier":"Compute RT-TDDFTB absorption spectra for sodium clusters Na13 to Na561 with a long-range corrected or optimally screened DFTB parameterization and compare the plasmon peak trend with the nonlocal hydrodynamic Drude model: if the predicted wavelength still redshifts with increasing size in the deep UV rather than blueshifting in the visible, the method's claim to general quantum plasmonics is contradicted.","tokens_in":37252,"feed_emoji":"🔬","tokens_out":7477,"duration_ms":70382,"temperature":0.7,"pith_summary":"The paper argues that real-time time-dependent density functional tight-binding (RT-TDDFTB) is a tractable quantum method for plasmonic nanostructures that full TDDFT cannot reach: systems of hundreds to thousands of atoms over picosecond timescales. The authors show it reproduces size-dependent plasmon shifts in noble metals, captures plasmon coupling in nanoparticle assemblies, describes second- and third-harmonic generation in gold nanorod dimers, and models hot-carrier-driven H2 dissociation on Ag, Au, Mg, and Al particles. This matters because it opens a purely quantum route to experimentally relevant particle sizes and timescales that were previously accessible only through classical or jellium models. The paper also states the method's limits: alkali-metal spill-out and long-range charge-transfer tunneling are not captured by current DFTB parameterizations.","feed_headline":"Tight-binding method scales quantum plasmonics to thousands of atoms","feed_subtitle":"Real-time DFTB captures size shifts, dimer coupling, harmonic generation, and H2 splitting at picosecond scales.","key_machinery":"The central object is the real-time time-dependent density functional tight-binding (RT-TDDFTB) equation of motion: the density matrix $\\rho_{\\mu\\nu}$ is propagated according to the Liouville-von Neumann form $\\partial\\rho/\\partial t = -i(S^{-1}H\\rho - \\rho H S^{-1}) - (S^{-1}D\\rho + \\rho D^{\\dagger} S^{-1})$, where $H$ is the time-dependent DFTB Hamiltonian, $S$ the overlap matrix, and $D$ the nonadiabatic coupling matrix. The Hamiltonian itself comes from a third-order expansion of Kohn-Sham DFT around a reference density, with orbital, self-consistent charge, and repulsive contributions parameterized in pretabulated Slater-Koster files. This machinery replaces the $O(N^{3-6})$ scaling of high-level quantum chemistry with a minimal-basis tight-binding problem, which is what lets the method propagate hundreds to thousands of atoms for picoseconds.","core_discovery":"On its own terms, the paper establishes that RT-TDDFTB, despite being a highly parameterized semiempirical method built on a minimal basis and Slater-Koster integrals, can describe the collective excited-state dynamics that define plasmonics. Density matrix propagation under the Liouville-von Neumann equation with Ehrenfest nuclear forces yields absorption spectra in which silver clusters show the anomalous blueshift with decreasing size caused by d-electron screening, whereas Mie theory does not; periodic arrays of Ag nanocubes show plasmon coupling redshifts and quenching with gap size; asymmetric Au nanorod dimers generate size- and polarization-dependent second and third harmonics; and H2 dissociation probabilities rise with particle size, with plasmon versus interband excitation differing between Ag and Au. The authors present these as evidence that a low-level quantum theory can reconcile classical electrodynamics and quantum mechanics for nanoplasmonics. The paper explicitly acknowledges that current DFTB fails to capture s-electron spill-out in alkali metal clusters and underestimates excited-state charge transport across sub-nanometer junctions, and that Ehrenfest dynamics limits electronic relaxation to the first tens of femtoseconds.","pith_inferences":["Pith inference: If long-range corrected or screened range-separated kernels were parameterized for metallic systems, the same RT-TDDFTB machinery could in principle fix the alkali-metal spill-out failure and the underestimated nanogap tunneling, turning the paper's stated limitations into testable extensions.","Pith inference: Because Ehrenfest dynamics lacks detailed balance and full electron-electron scattering, the picosecond photocatalytic predictions are most reliable for early-time hot-carrier transfer; coupling RT-TDDFTB to surface-hopping or Boltzmann transport schemes would likely change quantitative dissociation yields.","Pith inference: The demonstrated tractability suggests DFTB could serve as a screening tool for earth-abundant plasmonic metals such as Mg, Al, and Na for photocatalysis, provided new Slater-Koster parameter sets are developed; the paper leaves that parameter development as future work.","Pith inference: The method's reach into polaritonic chemistry in nanoparticle dimers and aggregates is implied but not developed in the paper; using RT-TDDFTB for cavity-modified chemistry would be a natural next step."],"forward_implications":["Plasmon absorption spectra can be computed quantum mechanically for particles up to roughly 1400 atoms, letting the excitonic-to-plasmonic crossover be mapped directly rather than extrapolated from small clusters.","For noble metals, the size-dependent blueshift in small Ag particles and the distinct Au plasmon evolution emerge naturally from d-electron screening, so DFTB offers a quantum check on classical dielectric models.","In nanoparticle aggregates, DFTB predicts plasmon coupling redshifts and oscillator-strength growth as gaps shrink, with quenching at sub-nanometer separations, giving a quantum boundary for classical electrodynamics.","RT-TDDFTB can compute nonlinear optical responses: harmonic generation in asymmetric Au nanorod dimers yields hyperpolarizabilities that grow with particle size and follow power-law field-intensity scaling.","For plasmon-driven catalysis, the dissociation threshold intensity decreases with particle size and depends on whether the drive is resonant with the plasmon or with interband transitions, giving concrete predictions for Ag, Au, Mg, and Al photocatalysts."],"supporting_citations":[{"why":"Introduces the original DFTB formulation that the paper builds on.","marker":"[37]"},{"why":"Provides the self-consistent-charge DFTB method (DFTB2) used for metallic and ionic systems.","marker":"[40]"},{"why":"Gives the tight-binding time-dependent response theory underlying TDDFTB.","marker":"[41]"},{"why":"Demonstrates DFTB for quantum plasmonics and benchmarks its speed against DFT.","marker":"[43]"},{"why":"Supplies the RT-TDDFTB implementation with semiclassical Ehrenfest electron-nuclear dynamics used throughout.","marker":"[71]"},{"why":"Reports the H2 photodissociation study on Ag and Au nanoparticles that anchors the photocatalysis section.","marker":"[85]"},{"why":"Applies RT-TDDFTB to aluminum nanoparticles and shows the role of d orbitals and oxidation.","marker":"[112]"},{"why":"Shows Mg nanoclusters generate hot carriers that drive H2 dissociation, used in the main-group metal examples.","marker":"[114]"},{"why":"Computes hot-carrier generation in gold and silver nanoclusters, supporting size-dependent plasmonic behavior.","marker":"[155]"},{"why":"Reports the RT-TDDFTB harmonic generation study in Au nanorod dimers with hyperpolarizabilities.","marker":"[181]"}],"fun_headline_variants":["Tight-binding method scales quantum plasmonics to thousands of atoms","DFTB makes quantum plasmonics practical for large systems","Quantum plasmonics at picosecond scale via tight-binding","Real-time tight-binding reveals quantum plasmonic effects","Scalable quantum plasmonics with density functional tight-binding"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that tight-binding parameters fitted to ground-state density functional theory can faithfully represent the delocalized excited-state transitions that make up a plasmon, an assumption the paper itself shows breaks down for alkali-metal spill-out and long-range charge transfer across nanogaps.","fun_headline_variants_meta":{"raw":{"variants":["Tight-binding method scales quantum plasmonics to thousands of atoms","DFTB makes quantum plasmonics practical for large systems","Quantum plasmonics at picosecond scale via tight-binding","Real-time tight-binding reveals quantum plasmonic effects","Scalable quantum plasmonics with density functional tight-binding"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000397,"raw_usage":{"total_tokens":2092,"prompt_tokens":973,"completion_tokens":1119,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":589,"completion_tokens_details":{"reasoning_tokens":1042}},"tokens_in":589,"tokens_out":1119,"duration_ms":10549,"temperature":1.0,"reasoning_tokens":1042,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:32:36.068181+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute RT-TDDFTB absorption spectra for sodium clusters Na13 to Na561 with a long-range corrected or optimally screened DFTB parameterization and compare the plasmon peak trend with the nonlocal hydrodynamic Drude model: if the predicted wavelength still redshifts with increasing size in the deep UV rather than blueshifting in the visible, the method's claim to general quantum plasmonics is contradicted.","supporting_citations":[{"cited_title":"A.; Suhai, S.; Della Sala, F.; Lugli, P.; Elstner, M.; Seifert, G.; Frauenheim, T","cited_arxiv_id":null,"evidence_quote":"Gives the tight-binding time-dependent response theory underlying TDDFTB."},{"cited_title":"Density Functional Tight Binding for Quantum Plasmonics","cited_arxiv_id":null,"evidence_quote":"Demonstrates DFTB for quantum plasmonics and benchmarks its speed against DFT."},{"cited_title":"K.; Schatz, G","cited_arxiv_id":null,"evidence_quote":"Reports the H2 photodissociation study on Ag and Au nanoparticles that anchors the photocatalysis section."},{"cited_title":"A.; Soldano, G","cited_arxiv_id":null,"evidence_quote":"Applies RT-TDDFTB to aluminum nanoparticles and shows the role of d orbitals and oxidation."},{"cited_title":"A.; Box, C","cited_arxiv_id":null,"evidence_quote":"Shows Mg nanoclusters generate hot carriers that drive H2 dissociation, used in the main-group metal examples."},{"cited_title":"A.; Berdakin, M.; Frauenheim, T.; Sánchez, C","cited_arxiv_id":null,"evidence_quote":"Computes hot-carrier generation in gold and silver nanoclusters, supporting size-dependent plasmonic behavior."},{"cited_title":"K.; Schatz, G","cited_arxiv_id":null,"evidence_quote":"Reports the RT-TDDFTB harmonic generation study in Au nanorod dimers with hyperpolarizabilities."}],"review_version":1}