{"id":"89fe5b00-8db7-44e3-a333-445c785bdd99","arxiv_id":"2602.14875","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"THz-STM can spectroscopically resolve Floquet-induced bulk band gaps and chiral edge states in graphene, and its signals can reveal edge-state dispersions and chirality.","lead":"This theory paper predicts that ultrafast terahertz scanning tunneling microscopy (THz-STM) can image the Floquet (light-induced) topological states in graphene with atomic-scale spatial and energy resolution. It provides a calculational framework and concrete simulated signals for gaps, edge states, interference patterns, and chirality that experimenters can look for.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'direct detection' claim assumes the circular pump is homogeneous in the sample and absent from the tunnel barrier; if junction dressing or near-field gradients contribute, the measured dips may not be the sample Floquet LDOS.","rationale":"I read the central claim as a proposal for a new probe: for it to hold, the measured rectified current must reflect the sample's Floquet LDOS. The formal derivation is internally consistent—Eq. (19) follows from the WBLA Meir-Wingreen formula with the pump present only in the sample Hamiltonian. The weakest point is not the mathematics but the mapping to experiment. I considered the alternative that the edge-state simulations use unrealistic drive parameters (E0 = 10 MV/cm, ħΩ = 1.5 eV) and that the ribbon-width breakdown scale may not transfer to realistic fields. That is a real limitation, but it is less load-bearing because the existence of chiral edge states is a robust topological consequence and the authors argue that real micrometer-scale samples provide even better localization at weaker drive. The junction-dressing assumption, by contrast, threatens the very observable used to infer all Floquet signatures, and the authors themselves identify it as a 'central challenge.' No code is shipped, which affects reproducibility but not the central claim's logic. Since the concern is substantive but explicitly acknowledged and potentially addressable by a concrete calculation, the appropriate verdict remains CONDITIONAL, matching the reader's assessment.","tokens_in":23255,"tokens_out":5091,"duration_ms":55775,"concrete_test":"Extend the real-time model to include a minimal junction-dressing term: add to the tip phase Φ in Eq. (14) a periodic contribution V_pump(t) = A·cos(Ωt + φ) with amplitude A from 0.1% to 10% of the THz bias amplitude, and recompute the Fig. 2b (and Fig. 3c) observables at the paper's parameters. If the dQ_rect/dV_pk minimum at eV_pk ≈ ħΩ/2 shifts by more than its width or acquires sideband structure at ±Ω, the measured signal is not cleanly the sample Floquet LDOS and the 'direct detection' claim requires conditioning; if the dip is robust for A ≤ 10%, the core assumption is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equation (19) expresses the measured cycle-averaged current as a convolution of sample Floquet LDOS, with the pump entering only through the Peierls-substituted sample Hamiltonian. This rests on the Sec. II A assumption that the circular pump is a homogeneous in-plane plane wave in the sample and is zero in the vacuum gap. In an actual STM junction the tip is present during the pump; it produces near-field gradients, polarization rotation, and out-of-plane components, and the optical field can modulate the tunneling barrier itself (photon-assisted/Tien-Gordon tunneling). None of these terms appear in the tip self-energy, Eq. (14), which contains only the THz bias phase. If such junction-dressing effects generate current sidebands at ±Ω with weight comparable to the sample LDOS features, the dips in dQ_rect/dV_pk at eV ≈ ħΩ/2 and the edge-state peaks in Fig. 3 could be partly or wholly junction artifacts. The authors explicitly flag this as a 'central challenge' in Sec. IV, but do not quantify it. The proposal's headline claim—that THz-STM gives 'direct local detection' of Floquet gaps—therefore depends on this separation being experimentally achievable, which is currently unverified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes THz-STM as a real-space, energy-resolved probe of Floquet topological states in graphene. The authors derive a nonequilibrium Green's-function tunneling formalism for time-periodic and pulsed drives, reducing the cycle-averaged current in the Floquet steady state to a convolution of the sample Floquet LDOS [Eq. (19)]. They then apply this formalism to bulk graphene and zigzag nanoribbons, predicting that dynamical Floquet gaps appear as dips in dQ_rect/dV_pk, that chiral edge states can be imaged at ribbon boundaries, that Floquet quasiparticle interference can reconstruct edge-mode dispersions, and that a chiral impurity produces a helicity-dependent LDOS signature. The paper is clearly written, the derivations follow standard Meir-Wingreen/NEGF methods, and the numerical parameters are stated transparently, including a candid admission that some ribbon calculations use driving parameters that are not experimentally realistic.","tokens_in":23524,"tokens_out":6805,"duration_ms":78003,"significance":"If the predictions are robust, the paper offers a genuinely new route to nanoscale, energy-resolved probing of light-induced topological states, complementing trARPES and transport. The formal reduction to a Floquet-LDOS expression is useful and likely to be adopted by others. The paper is also honest about its main limitations, which is a strength. However, the headline claim of 'direct local detection' depends on an idealized separation between the optical pump and the STM junction that is acknowledged but not quantified, and the ribbon-width results are obtained for strongly exaggerated gap parameters. These issues do not invalidate the formalism, but they do mean that the paper currently reads as a proof-of-principle proposal rather than an experimentally calibrated prediction.","major_comments":[{"comment":"The nanoribbon simulations use E0 = 10 MV/cm and ℏΩ = 1.5 eV, which the authors themselves state cannot be achieved in realistic experimental settings. The claim that edge-state protection breaks down below N = 10 is therefore parameter-specific. Since the edge-state decay length scales as ξ ~ 2ℏv_F/Δ, the crossover ribbon width for a realistic Floquet gap of order 10 meV would be orders of magnitude larger than the N = 10 value shown in Fig. 3(c). I do not object to proof-of-principle parameters, but the manuscript should either provide the scaling of the breakdown width with Δ, or clearly state that the N = 10 threshold is an artifact of the numerical parameters and not a prediction for experiments.","section":"§III B, Fig. 3"},{"comment":"The proposed 'smoking-gun' chirality probe uses a Haldane-like complex next-nearest-neighbor hopping iγ' with γ' = 0.7 eV localized to a single edge-adjacent hexagon. This is an ad hoc impurity model, and the chosen coupling is comparable to the artificially large Floquet gap used in the ribbon calculations. The predicted circular dichroism in the LDOS could depend sensitively on the impurity strength, spatial extent, and exact realization (magnetic adatom versus strain-induced gauge flux versus Haldane mass patch). A sensitivity analysis with respect to γ' and impurity position is needed before this can be presented as a robust experimental signature.","section":"§III D, Fig. 5"}],"minor_comments":[{"comment":"The T-matrix expression in Eq. (B4) is notationally unclear: the objects v, v∞, and the projector |0⟩⟨0| have different dimensionalities, and the final equality appears to conflate a scalar defect strength with the full Floquet-replica structure. Please define all quantities carefully.","section":"Appendix B, Eq. (B4)"},{"comment":"The definition of V_pk as max[V_DC + V_THz(t)] depends on the carrier-envelope phase φ_CEP. The text states φ_CEP = 2π/3 but then describes the amplitude as V0. Please clarify how V_pk is computed for a non-zero CEP and how it is related to the plotted horizontal axis.","section":"§III A, Eq. (27)"},{"comment":"The blue curves in Fig. 3(b) are described as 'the THz probe pulse with variable amplitude', but the vertical axis of the quasienergy spectrum is in energy units. The caption should explain that the bias waveforms are overlaid schematically and not to scale.","section":"Fig. 3 caption"},{"comment":"The term 'Floquet quasiparticle interference' is used for standing-wave patterns generated by hard-wall backscattering in a narrow ribbon, which differs from conventional impurity-induced QPI. Please define the term explicitly at first use and justify why the same name is appropriate.","section":"§III C"}],"recommendation":"major_revision","confidential_remarks":"The main stress-test concern about junction dressing is legitimate and lands on the central claim. The manuscript is transparent about the idealization, but the abstract and conclusion currently overstate what is predicted. The ribbon-width and chiral-impurity calculations also rely on model parameters that are not experimentally calibrated. I recommend major revision rather than reject because the NEGF/Floquet formalism itself is sound and no circularity was found; the required changes are quantitative qualification or additional modeling, not a restart."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper is a theory proposal, and the strongest thing you can say for it is that it knows what it is. The authors derive a Floquet version of the Meir-Wingreen formula for STM tunneling, get a clean expression for the cycle-averaged current as a Floquet LDOS convolution, and then use it to make concrete predictions: bulk gap dips in dQ/dV_pk, edge-state conductance peaks in ribbons, Floquet QPI maps, and helicity-dependent LDOS around a chiral impurity. None of those specific calculations appear in the cited literature, which is the paper's real contribution. The machinery is adapted from earlier NEGF work, but the application to THz-STM of Floquet topological matter is new and well executed. The numerics are internally consistent, the parameters are stated, and the appendix gives enough detail to reimplement.\n\nThe soft spots are real but mostly flagged by the authors themselves. The main one is the junction model: the circular pump is treated as a homogeneous in-plane field in the sample and zero in the vacuum gap, with tip near-field effects and photon-assisted tunneling neglected. The stress-test note is right that if junction dressing contributes comparable sidebands, the measured dips might not be the sample Floquet LDOS. But the authors say this in Sec. II A and return to it in Sec. IV as a central challenge. So it is not a hidden flaw; it is the known gap between the minimal model and a real STM junction. For a proposal paper that is acceptable, but it does mean the 'direct local detection' claim in the abstract is conditional, and I agree with the reader's CONDITIONAL verdict.\n\nThe second soft spot is that the edge-state and chirality simulations use drive parameters the authors call unrealistic (10 MV/cm, 1.5 eV), chosen because realistic parameters would need hundreds of lattice constants to resolve edge states. The bulk gap results use more realistic 350 kV/cm, 0.4 eV and those are the quantitative predictions most likely to be tested first. The ribbon-width breakdown scale, in particular, is inferred at unrealistic fields and may not carry over directly.\n\nNo code or data is shipped, but the method is described well enough that an independent implementation is feasible. I don't see that as a major problem for this type of paper.\n\nBottom line: this is a good, honest feasibility study with concrete falsifiable predictions. It deserves a serious referee and likely publication after revision. I'd cite it if I worked on Floquet topological matter or THz-STM. The main thing to check in review is whether the idealized junction can be made realistic enough for the bulk gap measurement, because that is the load-bearing prediction.","headline":"Sound theory proposal for THz-STM as a Floquet probe; the predictions are concrete and the caveats are acknowledged, but the 'direct detection' claim is conditional on the idealized junction.","tokens_in":24105,"tokens_out":2570,"would_cite":true,"duration_ms":25896,"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 terahertz scanning tunneling microscope can directly image the light-induced topological gaps and chiral edge states of driven graphene at atomic scale.","keywords":["THz-STM","Floquet topological insulator","graphene","local density of states","chiral edge states","nonequilibrium Green's functions","quasiparticle interference","time-resolved tunneling spectroscopy"],"falsifier":"A THz-STM experiment on undriven graphene, or on graphene driven with linearly polarized light at the same intensity, should show no conductance dip at half the photon energy; the Floquet interpretation predicts that dip only for circularly polarized drive. Equally, reversing the pump helicity should move the edge-state conductance peaks and reverse the impurity-induced LDOS asymmetry. If instead the same spectral features appear for all polarizations, the signal is a junction artifact rather than a Floquet gap.","tokens_in":23048,"feed_emoji":"🔬","tokens_out":4530,"duration_ms":42378,"temperature":0.7,"pith_summary":"This paper proposes ultrafast THz-STM as a real-space, energy-resolved probe of Floquet topological states in graphene. The authors derive a nonequilibrium tunneling formula in which the cycle-averaged rectified current reduces to a convolution of the Floquet local density of states with a Fermi function, so the measured dQ/dV spectrum directly maps the Floquet LDOS. Applied to bulk graphene, this resolves the dynamical gaps opened by circularly polarized drive; applied to nanoribbons, it images chiral edge states and shows the width at which edge protection breaks down. The point matters because existing probes average over large areas and cannot see nanoscale inhomogeneity or local edge modes in driven materials.","feed_headline":"THz-STM reveals light-made gaps and edge states in graphene","feed_subtitle":"A new tunneling formula maps rectified current to Floquet density of states at atomic scale.","key_machinery":"The workhorse is a nonequilibrium Green's-function formalism for time-dependent tunneling, in which the tip self-energy is local in space and carries the THz bias as a time-dependent phase, and the substrate acts as a thermal reservoir in the wide-band limit. In the Floquet steady state, the cycle-averaged current is expressed in a Floquet replica space as a convolution of the occupied and total Floquet LDOS at the tip position; this is the identity that turns a THz-STM rectified current into a local, energy-resolved Floquet spectrum. Around it sits the minimal-coupling substitution of the vector potential into the graphene hopping phases, which produces a circularly-polarized-drive mass ter","core_discovery":"The central claim is that a THz-STM junction, driven by a THz bias pulse while the sample is dressed by a circular pump, records a rectified charge whose derivative with respect to peak bias reproduces the Floquet local density of states. In the steady-state limit the cycle-averaged current is exactly a convolution of the occupied Floquet LDOS and the full Floquet LDOS, with the tip acting as a local, energy-resolved filter at a single lattice site. The authors demonstrate that this gives a plateau and a conductance dip at the Floquet hybridization gap, edge-localized conductance peaks in ribbons down to about thirty unit cells wide, and Fourier-visible scattering patterns that reconstruct t","pith_inferences":["If the local probe works as claimed, it should reveal nanoscale patches of topological and trivial Floquet regions in inhomogeneous samples, since the signal is site-specific; the authors note such patches are expected but do not simulate them.","The rectified-charge readout effectively integrates over a bias window set by the THz pulse; comparing different pulse shapes could separate instantaneous-bias artifacts from true Floquet LDOS, a testable extension the paper only sketches.","The gap signatures could be used as a quantitative pump-helicity detector: reversing circular polarization should flip the sign of the edge-state dichroism, providing a direct check.","Extending to cavity-modified environments, as the paper's discussion anticipates, would let the nanogap act as a tunable electromagnetic reservoir that shapes the Floquet dressing itself."],"forward_implications":["THz-STM can resolve the Floquet bulk gap locally, in real space, rather than averaged over macroscopic areas.","Chiral Floquet edge states appear as conductance peaks at ribbon edges and vanish when the ribbon narrows below the localization length, giving a concrete width scale for edge protection.","Floquet quasiparticle interference maps reconstruct the edge-state band dispersion and can image the absence of backscattering.","A chiral impurity creates a helicity-dependent LDOS enhancement or suppression that identifies the chirality of the edge mode.","The tunneling formalism applies to any periodically driven sample, not only graphene, so the same protocol can probe other driven quantum materials."],"fun_headline_variants":["THz-STM sees light-induced gaps and edge states in graphene","Atomic-scale probe of Floquet topology via THz-STM","Terahertz STM maps light-made band gaps and edge modes","Ultrafast THz-STM exposes Floquet states at atomic scale","New THz-STM technique reveals graphene's light-induced topology"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The predictions assume the pump can be treated as a spatially homogeneous in-plane plane wave inside the sample with no field in the tunneling gap, so that tip-induced near-field gradients, out-of-plane pump components, and photon-assisted junction dressing do not distort the measured current.","fun_headline_variants_meta":{"raw":{"variants":["THz-STM sees light-induced gaps and edge states in graphene","Atomic-scale probe of Floquet topology via THz-STM","Terahertz STM maps light-made band gaps and edge modes","Ultrafast THz-STM exposes Floquet states at atomic scale","New THz-STM technique reveals graphene's light-induced topology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000229,"raw_usage":{"total_tokens":1348,"prompt_tokens":812,"completion_tokens":536,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":556,"completion_tokens_details":{"reasoning_tokens":444}},"tokens_in":556,"tokens_out":536,"duration_ms":6264,"temperature":1.0,"reasoning_tokens":444,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T23:01:07.247236+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A THz-STM experiment on undriven graphene, or on graphene driven with linearly polarized light at the same intensity, should show no conductance dip at half the photon energy; the Floquet interpretation predicts that dip only for circularly polarized drive. Equally, reversing the pump helicity should move the edge-state conductance peaks and reverse the impurity-induced LDOS asymmetry. If instead the same spectral features appear for all polarizations, the signal is a junction artifact rather than a Floquet gap.","supporting_citations":[],"review_version":1}