{"id":"04924b5b-f725-41e5-9ffc-27257231c5a9","arxiv_id":"2509.10111","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A new DFTB+Maxwell framework simulates strong light-matter coupling in cavities efficiently, with demonstrations of 2D spectra, local molecular resolution, and cavity design.","lead":"This paper presents a fast simulation framework that couples molecules described by density-functional tight binding with a solution of Maxwell's equations for light in optical cavities. It demonstrates the method with three examples: two-dimensional spectra, molecule-resolved strong coupling, and inverse-designed cavities.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The independent-molecule scaling in Eq. (8) is untested at realistic packing densities, leaving the core 'realistic cavities' claim resting on an unvalidated assumption.","rationale":"The reader's conditional verdict identifies the same load-bearing approximation that I find most critical: the independent-molecule scaling of Eq. (8). The paper is transparent about this limitation, which is a point in its favor, but transparency does not reduce its load-bearing status. The abstract's promises of 'realistic cavities' and 'realistic chemical parameter spaces' are only valid if the no-overlap assumption holds at realistic molecular densities, and the three examples deliberately avoid that regime, so the key assumption is never tested. I considered the 2D-spectroscopy mismatch (Bloch emitters rather than DFTB molecules) and the absence of benchmarks and timings; these are real gaps, but they concern the completeness of the demonstration rather than the core physical validity of the method. The no-overlap assumption, if wrong, would invalidate the method's central output in its primary intended regime. The proposed dense-monolayer benchmark against an explicit-interaction reference would settle this. If the benchmark passes, the conditional acceptance is fully justified; if it fails, the framework would need an embedding or explicit many-molecule quantum region to support the original claim. Hence the reader's CONDITIONAL verdict remains unchanged.","tokens_in":18208,"tokens_out":8489,"duration_ms":75905,"concrete_test":"Simulate a dense pentacene monolayer (intermolecular spacing ~4 Å, matching experimental packing) inside a Fabry-Perot cavity with Drude mirrors. Compute the transmission Rabi splitting and per-molecule Fourier-resolved dipole spectra with DFTB+Maxwell, choosing N_M to match the experimental density. Repeat the same physical setup with all molecules placed in a single periodic DFTB (or DFTB-based QEDFT/Ehrenfest) cell so that exchange, charge transfer, and van der Waals interactions are treated explicitly, coupled to the same cavity mode(s). If the collective Rabi splitting or the local dipole spectra differ by more than ~10% between the two approaches, the independent-molecule scaling of Eq. (8) is invalid for dense realistic ensembles, and the central claim is overbroad.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim promises self-consistent treatment of large polaritonic ensembles in realistic cavities with access to both global and local properties. The method's matter model, introduced in Sec. 2 via Eq. (8), assigns one isolated DFTB molecule (or a scaled copy) to each Maxwell grid point, with P = N_M <mu>. The paper explicitly states: 'our current assignment of molecules to the Maxwell grid, as well as the scaling argument of Eq. (8) assumes no overlap of the electronic structure between different Maxwell-grid points. Thus any intermolecular interactions remain entirely classical (no exchange and no correlation effects).' This assumption is load-bearing because in dense molecular ensembles typical of cavity-modified chemistry (organic thin films, molecular liquids, J-aggregates), intermolecular distances are 3-10 Å, where wavefunction overlap, charge transfer, and van der Waals/exciton coupling materially affect collective Rabi splittings and local molecular response. The three illustrative applications do not exercise this regime: N2 molecules in Sec. 3.2 are placed on a grid with ~1 nm spacing, and pentacene groups in Sec. 3.3 are separated by 300 nm, so the assumption is satisfied by construction. The abstract claims access to 'realistic chemical parameter spaces' and 'large polaritonic ensembles in realistic cavities,' but the presented calculations only validate the dilute or spatially separated limit. If the independent-molecule approximation fails at realistic packing densities, the advertised global and local properties would be quantitatively wrong, directly undermining the central claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a multiscale computational framework (DFTB+Maxwell) that couples real-time density-functional tight binding (DFTB) for molecular electronic dynamics with finite-difference time-domain (FDTD) solutions of Maxwell's equations. The stated goal is to enable non-perturbative simulations of strong light-matter coupling in cavities with both collective (global) and local, molecule-resolved information. The paper reports three illustrative applications: a two-dimensional spectroscopy simulation for a strongly coupled cavity-emitter system, collective strong coupling of N2 molecules in a one-dimensional Fabry-Perot cavity with a Rabi-splitting analysis, and inverse design of two-dimensional cavities optimized for coupling to pentacene molecules. The authors claim that the method is computationally efficient and allows near real-time exploration of chemical parameter spaces on standard hardware.","tokens_in":18604,"tokens_out":4690,"duration_ms":39902,"significance":"If the framework performs as claimed, it would provide a practical bridge between quantum-chemistry descriptions of molecules and the spatially resolved electromagnetic environment of realistic cavities, an area where current ab initio QED tools are restricted to few molecules or effective single-mode treatments. The conceptual direction is sound: the Maxwell equations are standard, the rt-TD-DFTB formulation is established, and the explicit statement of the independent-molecule approximation in Eq. (8) is a commendable disclosure of a key limitation. However, the paper's demonstrations do not yet substantiate the advertised predictive capabilities: the two-dimensional spectroscopy example uses Bloch equations rather than the DFTB+Maxwell method, the Drude mirror parameters are explicitly fictitious, and no benchmark against analytic models, exact QED calculations, or experimental spectra is provided. The strength of the paper lies in the originality of the coupled framework and its potential user accessibility, but quantitative validation is currently missing.","major_comments":[{"comment":"The two-dimensional spectroscopy demonstration is not performed with the DFTB+Maxwell method described in Sec. 2. The text states that the molecular layer consists of 'two-level emitters treated using standard Bloch equations,' and the resulting spectrum in Fig. 3c reports a Rabi splitting of 242 meV. This does not demonstrate the claimed capability of the DFTB+Maxwell framework. Yet Sec. 5 summarizes the paper as offering 'non-perturbative, quantitative insights into (multi-dimensional) spectroscopic observables (see e.g. Sec. 3.1).' The central claim for 2D spectroscopy therefore rests on an example that is external to the method. Please either provide a DFTB-based 2D spectrum or explicitly reframe the claim as a planned capability rather than a demonstrated one.","section":"Sec. 3.1, Fig. 3"},{"comment":"The independent-molecule scaling P = N_M ⟨μ⟩ with the stated assumption 'no overlap of the electronic structure between different Maxwell-grid points' is load-bearing for the abstract's claim of access to 'realistic chemical parameter spaces' and 'large polaritonic ensembles in realistic cavities.' However, the presented applications are dilute by construction: N2 molecules in Sec. 3.2 are spaced at approximately 1 nm grid spacing, and pentacene groups in Sec. 3.3 are separated by 300 nm. No test at realistic molecular packing densities is provided, and the Outlook explicitly lists 'Beyond Gaseous Phase' as a future direction. The claim that the framework addresses realistic condensed-phase ensembles is therefore not supported by the current evidence. Please either benchmark the independent-molecule approximation at relevant densities or restrict the claims to dilute ensembles.","section":"Sec. 2, Eq. (8)"},{"comment":"The quantitative predictions of the collective strong-coupling results are undermined by the explicit statement that the Drude parameters (Ω_p = 34 eV, γ = 0.181 eV) 'are fictitious and were selected to fit the specific frequencies of interest.' This means the Rabi splittings in Fig. 4 do not correspond to any real mirror material, and the dependence on the arbitrarily chosen parameters is not analyzed. Moreover, no comparison against an analytic coupled-mode model or a known experimental system is provided, so the accuracy of the method in a realistic cavity is untested. Please add a validation case with realistic material parameters and a quantitative benchmark.","section":"Sec. 3.2, Drude parameters"}],"minor_comments":[{"comment":"The manuscript contains a large unintegrated block of text beginning with '2 FIG. 1. 1D scheme of the numerical implementation.' and ending with 'integration time step', which duplicates a figure caption and method section from another publication. This material interrupts the flow of the paper and should be removed or properly integrated into the main text.","section":"After Sec. 3 heading, before Sec. 3.1"},{"comment":"There is a duplicated parenthetical: 'density-functional tight binding (density-functional tight binding (DFTB))' appears in the abstract and again in the introduction. This should be corrected to 'density-functional tight binding (DFTB)'.","section":"Abstract and Sec. 1"},{"comment":"The claim of 'almost real-time exploration of chemical or physical parameters on a desktop machine' is not supported by any timing data or hardware description. A brief runtime benchmark (e.g., wall-clock time for the N2 cavity simulation with 80 molecules) would make this claim concrete and verifiable.","section":"Sec. 5"},{"comment":"The first electronic transition of N2 is quoted as 13.902 eV based on DFTB. Given that this value is used to set the cavity resonance, the authors should compare it with an experimental or higher-level theoretical reference value to establish the accuracy of the DFTB description in this context.","section":"Sec. 3.2"},{"comment":"Several references are to unpublished work (Refs. 23 and 41 are in preparation or preprint) and the companion paper Ref. 33 is an arXiv preprint. Please clarify the status of these works so that readers can assess what is established versus what is still in development.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a perspective-style paper that introduces a potentially useful framework but currently claims more than it demonstrates. The separation between the method (DFTB+Maxwell) and the 2D-spectroscopy example (Bloch equations) is a substantive gap, and the fictitious Drude parameters and lack of benchmarking limit the quantitative value. The duplicated passage in Sec. 3 is a serious presentation defect that suggests the manuscript is not yet in journal-ready form. The paper also leans heavily on self-citations, and the relationship to Ref. 33 (the companion arXiv preprint) should be clarified: it is unclear what is new in this submission relative to that work. These issues are fixable, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Things to know first: this is a progress report from the authors' own DFTB+Maxwell program, not a stand-alone methods paper. The core coupling already appeared in their companion work (ref. 33); what's new here are three illustrative applications: a pulse-subtraction protocol for 2D spectroscopy, a position-resolved collective strong-coupling example that shows the sqrt(N) scaling breakdown, and inverse-designed 2D cavities for pentacene.\n\nIt does several things well. The ingredients are standard and the self-consistent Maxwell-matter feedback is genuinely non-perturbative. The sqrt(N) breakdown is a clean physics point: once molecules span more than a fraction of the mode wavelength, the uniform-field assumption behind the usual scaling fails. The molecule-resolved Fourier spectra show local mode structure and dark-state features that a transmission spectrum hides. The inverse-design exercise is a sensible proof of concept, and the authors are open about which parameters are fictitious.\n\nThe soft spots are proportionate but real. The headline 2D spectroscopy demo uses two-level Bloch emitters, not DFTB molecules; the pulse-subtraction idea is promising but the claimed DFTB+Maxwell 2D capability is not actually demonstrated. All parameters are illustrative: Drude mirrors are explicitly fictitious, and the N2 transition energy sits at 13.9 eV without any benchmark against exact QED, analytic models, or experiment. The independent-molecule assumption behind Eq. (8) is stated openly in the text, and all examples are dilute (nm-spaced N2; 300 nm-separated pentacene groups), so the 'realistic chemical parameter spaces' claim is, for now, a dilute-regime claim. The outlook acknowledges the embedding steps needed to go further.\n\nWho this is for: experimentalists and modelers in polaritonic chemistry who want a fast screening tool and an honest map of where this method stands. On that basis, it deserves a serious referee: the trajectory is credible, the examples illustrate real physics, and the limitations are in the paper rather than hidden. I would send it to review, but I'd push hard for benchmarks, realistic parameters, and preferably code/data release before it's cited as a validated method.","headline":"A readable, honest progress report on the authors' DFTB+Maxwell program: the new demonstrations are useful but the headline 2D-spectroscopy claim outruns the evidence, and the method is as-yet unvalidated outside the dilute-regime examples it actually shows.","tokens_in":19039,"tokens_out":3046,"would_cite":false,"duration_ms":25948,"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":"A combined DFTB+Maxwell method simulates strong light-matter coupling in realistic cavities non-perturbatively, resolving both collective and molecule-level effects.","keywords":["density-functional tight binding","finite-difference time-domain","strong light-matter coupling","polaritonic chemistry","two-dimensional spectroscopy","collective strong coupling","cavity inverse design","real-time electron dynamics"],"falsifier":"Compare DFTB+Maxwell transmission spectra and molecule-resolved dipole Fourier transforms for a dense molecular layer against a full minimal-coupling Maxwell-TDDFT calculation on the same geometry; if the Rabi splitting or local spectral features deviate once intermolecular distances shrink to where wavefunction overlap matters, the no-overlap grid assumption is the culprit. A simpler experiment is to keep the number of molecules fixed while compressing their spatial distribution; if the Rabi splitting changes with density beyond the field-profile effect, collective coupling depends on more than the cavity mode and the scaling argument of Eq. (8) fails.","tokens_in":18005,"feed_emoji":"⚛️","tokens_out":6028,"duration_ms":46366,"temperature":0.7,"pith_summary":"The paper introduces a computational method that couples density-functional tight binding (DFTB) for molecules to a finite-difference time-domain (FDTD) solution of Maxwell's equations, and claims this lets one study strong light-matter coupling in realistic optical cavities non-perturbatively for large molecular ensembles. The method feeds the cavity electric field back into each molecule's quantum-mechanical time evolution while the molecular dipoles drive the field, so collective and local effects are treated self-consistently. The authors demonstrate three capabilities: non-perturbative two-dimensional spectra, molecule-resolved information inside collectively coupled ensembles, and inverse-designed cavities that couple spatially separated molecular groups. Because a one-dimensional Maxwell grid with three-dimensional molecules runs on a desktop machine, the authors argue the tool enables near real-time exploration of chemical parameter spaces and can bridge experiment and theory in polaritonic chemistry.","feed_headline":"Cavity strong-coupling chemistry simulated in near real time","feed_subtitle":"A self-consistent DFTB+Maxwell loop resolves collective and local polariton effects that perturbative models miss.","key_machinery":"The load-bearing object is the self-consistent loop between the FDTD Maxwell solver, which updates the electric and magnetic fields from the polarization current, and the real-time time-dependent DFTB propagation of each molecule's density matrix through the Liouville-von Neumann equation in a non-orthogonal basis. The time-dependent DFTB Hamiltonian includes the electric field at the molecule's position under the dipole approximation, and the molecular polarization entering Maxwell's equations is written as a concentration-scaled expectation value of the molecular dipole, with a single simulated molecule representing many molecules at a grid point. For two-dimensional spectra, the scheme propagates several coupled Maxwell-Schrödinger systems in parallel: one with all three pulses, one with the first pump only, and one with the second pump only, then subtracts the single-pump fields from the three-pulse field to isolate the third-order nonlinear signal.","core_discovery":"The central claim is that the combined DFTB+Maxwell framework correctly captures the non-perturbative feedback between a macroscopic electromagnetic environment and the microscopic electronic structure of many molecules, without partitioning the system into a 'system' and a 'bath' or invoking perturbative, rotating-wave, or Markov approximations. Concretely, the paper shows that the collective Rabi splitting grows with the square root of the number of molecules only while the ensemble is smaller than roughly half the cavity-mode wavelength; for larger ensembles the spatial mode profile breaks this scaling. Fourier-transforming individual molecular dipoles reveals a position-dependent polariton response that traces the cavity mode, plus dark-state and off-resonant features invisible in the transmission spectrum. The paper further shows that topology-optimized cavities can be designed to maximize the field at chosen molecular locations, and that spatially separated molecular groups coupled by such a cavity form a single polaritonic ensemble, evidenced by growing Rabi splitting as groups are added.","pith_inferences":["A direct consequence the paper leaves implicit is that the method is on its surest footing for dilute or gas-phase ensembles; extending to condensed phases will require explicit solvation or embedding, as the paper itself notes.","The field-subtraction strategy for two-dimensional signals generalizes in principle to higher-order nonlinear responses by adding more auxiliary propagations, at linearly growing cost.","The demonstrated position dependence of the Rabi splitting suggests that experiments varying the location or thickness of a molecular layer inside a cavity should see the same mode-profile modulation, offering a direct quantitative test of the framework.","Because the electromagnetic field is treated classically, vacuum-field quantum effects such as modified London dispersion or spin-glass-like correlations lie outside the current framework; recovering them would need multi-trajectory Ehrenfest dynamics or transverse exchange-correlation functionals, directions the paper sketches."],"forward_implications":["Two-dimensional spectra of strongly coupled molecular ensembles can be computed without rotating-wave, Markov, or perturbative assumptions, with the field-subtraction scheme isolating third-order signals.","Molecule-resolved spectra expose dark states and locally varying polariton weights that ensemble transmission hides, giving a route to interpret position-dependent experimental observations.","The breakdown of square-root-of-N scaling for extended ensembles implies that the spatial distribution of molecules, not just their total number, controls collective light-matter coupling strength.","Inverse-designed cavities can deliberately couple molecular groups separated by hundreds of nanometers, enabling design of polaritonic networks with targeted connectivity.","For one-dimensional cavities the framework runs in near real time on a desktop, making cavity-modified chemistry parameter sweeps practical before experiments."],"supporting_citations":[{"why":"Introduces the original DFTB+Maxwell coupling scheme that this paper extends and demonstrates.","marker":"[33]"},{"why":"Supplies the self-consistent-charge DFTB Hamiltonian defining the molecular electronic-structure level.","marker":"[45]"},{"why":"Provides the real-time TD-DFTB density-matrix propagation used for the molecular response.","marker":"[48]"},{"why":"The DFTB+ package used as the library implementing the electronic-structure and dynamics engine.","marker":"[53]"},{"why":"The Ehrenfest-Maxwell-Pauli-Kohn-Sham framework that DFTB+Maxwell approximates for practical molecular ensembles.","marker":"[27]"},{"why":"Full minimal-coupling Maxwell-TDDFT, the more complete but computationally heavier benchmark this method approximates.","marker":"[32]"},{"why":"Drude-Lorentz parametrizations used for the metallic mirror response in the cavity simulations.","marker":"[42]"},{"why":"Supplies the inverse-design and topology-optimization methodology used to engineer the cavities.","marker":"[59]"},{"why":"Experimental evidence of position-dependent Rabi splitting that supports the local mode-profile results.","marker":"[58]"}],"fun_headline_variants":["Simulating strong light-matter coupling efficiently in cavities","Non-perturbative cavity QED simulation for molecular ensembles","DFTB+Maxwell: self-consistent strong-coupling simulations","Fast simulations of collective polariton effects in cavities","Cavity design meets quantum chemistry: DFTB+Maxwell"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The method places one DFTB molecule at each Maxwell grid point and multiplies its dipole by a concentration factor, assuming the electronic structures of molecules at different grid points do not overlap; all intermolecular interactions are therefore classical, with no exchange or correlation between molecules.","fun_headline_variants_meta":{"raw":{"variants":["Simulating strong light-matter coupling efficiently in cavities","Non-perturbative cavity QED simulation for molecular ensembles","DFTB+Maxwell: self-consistent strong-coupling simulations","Fast simulations of collective polariton effects in cavities","Cavity design meets quantum chemistry: DFTB+Maxwell"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000769,"raw_usage":{"total_tokens":3438,"prompt_tokens":1010,"completion_tokens":2428,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":626,"completion_tokens_details":{"reasoning_tokens":2345}},"tokens_in":626,"tokens_out":2428,"duration_ms":378936,"temperature":1.0,"reasoning_tokens":2345,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:56:47.140719+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare DFTB+Maxwell transmission spectra and molecule-resolved dipole Fourier transforms for a dense molecular layer against a full minimal-coupling Maxwell-TDDFT calculation on the same geometry; if the Rabi splitting or local spectral features deviate once intermolecular distances shrink to where wavefunction overlap matters, the no-overlap grid assumption is the culprit. A simpler experiment is to keep the number of molecules fixed while compressing their spatial distribution; if the Rabi splitting changes with density beyond the field-profile effect, collective coupling depends on more than the cavity mode and the scaling argument of Eq. (8) fails.","supporting_citations":[{"cited_title":"M.; Bonaf \\'e , F","cited_arxiv_id":null,"evidence_quote":"Introduces the original DFTB+Maxwell coupling scheme that this paper extends and demonstrates."},{"cited_title":"Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties","cited_arxiv_id":null,"evidence_quote":"Supplies the self-consistent-charge DFTB Hamiltonian defining the molecular electronic-structure level."},{"cited_title":"P.; Aradi, B.; Hourahine, B.; Medrano, C","cited_arxiv_id":null,"evidence_quote":"Provides the real-time TD-DFTB density-matrix propagation used for the molecular response."},{"cited_title":"The Journal of chemical physics 2020, 152, 124101","cited_arxiv_id":null,"evidence_quote":"The DFTB+ package used as the library implementing the electronic-structure and dynamics engine."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Ehrenfest-Maxwell-Pauli-Kohn-Sham framework that DFTB+Maxwell approximates for practical molecular ensembles."},{"cited_title":"P.; Albar, E","cited_arxiv_id":null,"evidence_quote":"Full minimal-coupling Maxwell-TDDFT, the more complete but computationally heavier benchmark this method approximates."},{"cited_title":"D.; Djuri s i \\'c , A","cited_arxiv_id":null,"evidence_quote":"Drude-Lorentz parametrizations used for the metallic mirror response in the cavity simulations."},{"cited_title":"Y.; Jin, W.; Vuckovi \\'c , J.; Rodriguez, A","cited_arxiv_id":null,"evidence_quote":"Supplies the inverse-design and topology-optimization methodology used to engineer the cavities."},{"cited_title":"A.; Genet, C.; Ebbesen, T","cited_arxiv_id":null,"evidence_quote":"Experimental evidence of position-dependent Rabi splitting that supports the local mode-profile results."}],"review_version":1}