{"id":"bde2ff64-391e-47a0-828c-c40d363d1606","arxiv_id":"2607.29501","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Rabi oscillations between cyclotron resonance and a Fabry-Perot cavity mode set in one cavity round-trip time after excitation, showing that Landau polaritons form dynamically.","lead":"Phase-resolved terahertz spectroscopy of a 2D electron gas in a Fabry-Perot cavity shows that the Rabi beating of Landau polaritons begins only after one cavity round-trip time, not immediately. The result illustrates that strong light-matter coupling builds up dynamically as the cavity standing wave forms.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The delayed-formation claim rests on a null result inside the first cavity round-trip, where the CR burst is explicitly weak; without noise characterization the tuned/detuned similarity in Fig. 1d may be a common noise floor, and the thin-sample 'first node' is extrapolated, not measured.","rationale":"The reader's conditional verdict is appropriate, and the weakest assumption identified is indeed the most load-bearing one. The central claim—that strong coupling is absent for one cavity round-trip—rests on a null observation in precisely the interval where the signal is weakest. The text says the CR response following the first pulse is weak, so the 'similarity' of tuned and detuned differential traces in Fig. 1d may simply reflect a common noise floor rather than a physical absence of coupling. The thin-sample support is even weaker: the first node is extrapolated from later nodes and cannot resolve the delay independently. The fitted transfer-matrix simulation is not an independent verification because it is adjusted to the frequency-domain data. All of these issues point to the same conclusion: the delayed-formation claim is plausible but not quantitatively established by the present data. I therefore agree with the reader and recommend keeping the CONDITIONAL verdict. The proposed concrete test—noise-floor characterization of the first-cycle window and a control measurement with the back mirror suppressed—would settle whether the first-cycle similarity is informative. If the similarity persists with the first burst well above the noise floor, the delayed-formation claim would be substantially strengthened.","tokens_in":9151,"tokens_out":19415,"duration_ms":227358,"concrete_test":"From the raw time-domain traces, compute the RMS noise in the t=0–9 ps window using the B=0 cross-polarized trace and/or repeated scans, then quantify the similarity between the tuned (B=1.48 T) and detuned (B=1.36 T) ∂R_x/∂B traces in that window with a normalized cross-correlation and its uncertainty. If the difference between the two traces is within the noise floor, the 'matter-only' first cycle is not established. Then repeat the thick-sample measurement with higher excitation intensity or longer averaging so that the first CR burst is well above the noise floor; if the tuned trace still shows no Rabi signature before 9 ps, the delayed-formation claim is supported. Optionally, measure a control sample with the back surface roughened to suppress cavity feedback and check that the first 9 ps response is unchanged while the later Rabi oscillations disappear.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Rabi oscillations begin only after one cavity round-trip, inferred from (i) the similarity of tuned and detuned ∂R_x/∂B traces in the first 9 ps of the thick sample (Fig. 1d) and (ii) an extrapolated first node at about 4 ps in the thin sample (Fig. 2a). The load-bearing assumption is that this first-cycle similarity reflects an absence of light–matter coupling. However, the manuscript explicitly states that the CR burst following the first pulse is weak ('Weak CR oscillation follows the 1st peak'), so the tuned and detuned traces in the t=0–9 ps window may both be sitting at the noise floor; their similarity would then be an uninformative null result rather than evidence of delayed coupling. No error bars, noise-floor estimate, or null-test control (e.g., a sample with the back mirror suppressed) is provided. For the thin sample, echoes dominate before 10 ps and the first node at 4 ps is predicted, not observed, so it cannot independently verify the delay. Since the positive Rabi-oscillation evidence appears only after the first echo, the delay conclusion is currently conditional on an unvalidated null.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports phase-resolved THz time-domain spectroscopy measurements of cyclotron resonance in a two-dimensional electron gas coupled to Fabry-Perot cavity modes. Using cross-polarized detection and magnetic-field differential traces, the authors observe temporal beating that they identify as Rabi oscillations of Landau polaritons. Their central claim is that these oscillations do not start immediately after the THz pulse excites the cyclotron resonance, but only after a delay equal to one cavity round-trip time, implying that the strong-coupling regime is dynamically established as the cavity mode field builds up. Evidence is presented for a thick (383 um) and a thin (131 um) substrate sample, supported by transfer-matrix simulations of the frequency- and time-domain response.","tokens_in":9610,"tokens_out":7059,"duration_ms":86888,"significance":"If the delayed-formation claim survives scrutiny, this would be a notable contribution: it would be a direct, phase-resolved observation of polariton Rabi oscillations in a Landau-polariton system and would give a concrete picture of the finite build-up time of a Fabry-Perot mode. The experimental approach—cross-polarized THz-TDS with magnetic-field differential detection—is well suited to isolating the cyclotron-resonance response, and the sign reversal of the cross-polarized signal under magnetic-field reversal convincingly identifies the cyclotron origin of that signal. The main limitation is that the delay conclusion is currently supported by a null result in a weak-signal time window and by an extrapolated node in the thin-sample data, so the strength of the claim is somewhat ahead of the evidence. The paper would be significantly strengthened by quantitative noise characterization and a direct test of the delayed-onset hypothesis.","major_comments":[{"comment":"The inference that the first cavity cycle (t = 0–9 ps) is free of strong coupling rests on the similarity between the tuned and detuned differential traces in a time window where the manuscript itself describes the cyclotron-resonance signal as weak ('Weak CR oscillation follows the 1st peak'). Without an estimate of the noise floor, error bars on the traces, or a null-test control—for example, a sample without the back mirror, or a simulation with the coupling artificially suppressed during the first round trip—the tuned/detuned similarity could simply be a common noise floor rather than evidence for delayed coupling. Please provide a quantitative measure of agreement in the t = 0–9 ps window (e.g., residual norm compared to the noise level) and a noise-floor estimate.","section":"Fig. 1d and accompanying text"},{"comment":"The thin-sample evidence for the delay is not independent. The 'first node at about 4 ps' in Fig. 2a is extrapolated from the Rabi period measured from the nodes observed after 10 ps, so it cannot independently verify the delay. Moreover, the numerical reasoning is not transparent: if the observed node spacing is 9 ps, a first node near T_Rabi/4 ≈ 4.5 ps is expected even if the coupling started at t = 0, so the conclusion that the data indicate a delay 'close to the cavity period of 3 ps' does not follow from the stated numbers. A full time-domain fit with the onset time as a free parameter, comparing models with and without a delayed onset, is needed to support the claim.","section":"Thin-sample prediction (final paragraphs before Summary)"},{"comment":"The transfer-matrix simulation is used to support the time-domain interpretation, but the parameters are fitted to the same frequency-domain spectra that establish the strong coupling. It is therefore important to state explicitly whether the delayed onset in the simulated time-domain traces is an input or an emergent output of the linear-response model. If it is an emergent output, the simulation should be shown to reproduce the first-cycle null without added assumptions; if the delay is imposed, the simulation cannot serve as independent confirmation. Please report the fitted parameter values with uncertainties and show the simulated traces on the same t = 0–10 ps scale as the data.","section":"Supplementary Sec. IIIC / time-domain simulation"}],"minor_comments":[{"comment":"The labels 'CR tunned' and 'CR detunned' should be 'tuned' and 'detuned'; also 'time scale tics' should be 'time scale ticks'.","section":"Fig. 1d"},{"comment":"The scaling formula 'The Rabi splitting scales as d p N fm/Vm [1]' contains a formatting artifact and should be typeset as a proper equation with definitions of the symbols.","section":"Text near Eq. (1)"},{"comment":"The electron density and mobility values are typeset with missing superscripts ('ns ∼3×10 11 cm−2' and 'µ >10 5 cm2/Vs'); please correct the exponent formatting.","section":"Sample parameters"},{"comment":"Reference [2] contains an unformatted URL line break in the citation text; please ensure the bibliographic entry is clean.","section":"References"},{"comment":"The color scale and position axes for the simulated electric-field distributions are not defined in the caption; please specify the color mapping and the meaning of the vertical axis.","section":"Fig. 2e/f"},{"comment":"The text repeatedly refers to Supplemental Material sections (Sec. IIB, IIIC, IV, V) without giving a preview of their content in the main text; please ensure the supplement is complete and archived.","section":"Supplemental material"}],"recommendation":"major_revision","confidential_remarks":"The delayed-formation result is the headline claim and it currently hinges on an uncharacterized null result in the thick sample and a circular extrapolation in the thin sample. I recommend requesting the noise analysis and a direct onset-time fit before considering the paper for publication. The topic and technique are appropriate for the journal, but the evidence needs to be brought up to the level of the claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the headline: this is a genuinely new phase-resolved time-domain observation of Rabi oscillations in Landau polaritons, and the claim that strong coupling takes one cavity round-trip to build up is a physically sensible interpretation. The paper is worth a serious referee, but as it stands the delay claim rests on a null result in the first cavity period that needs a proper noise analysis before it is quantitative.\n\nWhat the paper does well: the cross-polarized THz-TDS setup isolates cyclotron emission cleanly, the magnetic-field differential is a smart way to extract the field-dependent component, and the two samples with different substrate thicknesses give a consistent scaling of the Rabi period with cavity volume. The transfer-matrix model reproduces the frequency-domain Faraday angle maps and captures the time-domain beats qualitatively. That is solid experimental work.\n\nThe soft spot is the central claim of delayed onset. In the thick sample, the evidence that the first 9 ps show no coupling is the similarity of tuned and detuned differential traces. But the paper itself says the CR oscillation after the first pulse is weak, and without a noise floor or error bars that similarity could just be two traces sitting at the noise floor. The thin-sample first node at 4 ps is predicted from the measured Rabi period, not observed, so it doesn't independently verify the delay. The transfer-matrix simulation is fitted to the same spectra that define the strong coupling, so it can't settle the question either. None of this undermines the observation of the Rabi beat after the first echo; it just means the 'delayed formation' part is plausible but not demonstrated.\n\nFor the review: the authors should provide an estimate of the noise level in the first cavity period, ideally with a null-test control or a simulation with and without the back mirror. That would turn a suggestive observation into a convincing one. The paper deserves peer review; it is a new time-domain result in a field where most evidence is spectral splittings. I'd send it to a good referee, but with a note to focus on the first-cycle null.","headline":"A genuinely new time-domain observation of Landau polariton Rabi oscillations, but the delayed-onset claim rests on a null result in the first cavity cycle and needs a noise analysis before it is quantitative.","tokens_in":9989,"tokens_out":2270,"would_cite":true,"duration_ms":27414,"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":"Using phase-resolved terahertz spectroscopy, this paper shows that Landau-polariton Rabi oscillations begin only after a delay of one cavity round-trip time, not immediately after cyclotron-resonance excitation.","keywords":["Landau polaritons","cyclotron resonance","terahertz time-domain spectroscopy","Rabi oscillations","strong light-matter coupling","Fabry-Perot cavity","two-dimensional electron gas","Faraday rotation"],"falsifier":"A measurement of the thick sample with improved signal-to-noise that resolves a Rabi-beating node within the first 9 ps, near 5.5 ps for a 22 ps Rabi period, would disprove the one-round-trip delay.","tokens_in":9010,"feed_emoji":"🕐","tokens_out":6958,"duration_ms":64071,"temperature":0.7,"pith_summary":"The paper reports a time-domain measurement of light-matter coupling in a terahertz cavity. It claims that when cyclotron resonance of a two-dimensional electron gas is strongly coupled to Fabry-Perot cavity modes, the resulting Landau polaritons exhibit Rabi oscillations whose onset is delayed by one cavity round-trip time. For the first cavity period the cyclotron-resonance signal behaves as if the cavity were absent; the periodic energy exchange between matter and light appears only after the cavity field has built up. If true, this shows directly that the strong-coupling regime is not instantaneous but forms dynamically with the standing wave.","feed_headline":"Polariton Rabi oscillations begin one cavity round-trip late","feed_subtitle":"Phase-resolved THz traces show the light-matter beat appears only after the cavity field builds up.","key_machinery":"The load-bearing mechanism is the train of Fabry-Perot echoes: each THz pulse echo arriving at the two-dimensional electron gas excites cyclotron resonance, and the cavity round-trip time, equal to twice the time of flight through the substrate, sets the delay before the counter-propagating waves interfere into a standing mode. The Rabi beat period is extracted from $B$-field-differential traces $\\partial R_x/\\partial B$ in co-polarization and from nodes in cross-polarized $R_y$ traces, while the zigzag Faraday-angle pattern identifies the strong-coupling condition. The cavity round-trip time is thus the central timescale that determines when polaritonic beating can begin.","core_discovery":"The central discovery is that Landau polaritons, hybrid states formed by coupling the cyclotron resonance of a two-dimensional electron gas to Fabry-Perot cavity modes, show Rabi oscillations in phase-resolved THz reflection that begin only after a delay of roughly one cavity round trip. During the first cavity period, the cyclotron-resonance signal evolves as a matter-only excitation; the periodic energy exchange with the cavity field appears only after counter-propagating waves have interfered to form the standing cavity mode. The evidence comes from magnetic-field differential traces that are nearly identical for tuned and detuned fields in the first 9 ps of the thick sample, and from an inferred first node at about 4 ps in the thin sample with a 3 ps cavity period.","pith_inferences":["If the one-round-trip delay is a general rule, then shrinking the cavity should shorten the onset time in proportion; a series of substrate thicknesses would provide a direct test.","The delay also implies that ultrafast switching of polaritons cannot respond faster than the cavity build-up time; probes shorter than one round trip should see a matter-like response.","The extrapolated first node in the thin sample could be measured directly by suppressing the early echoes, for example with an anti-reflection coating, to confirm the delay without relying on inference."],"forward_implications":["In thick-substrate samples, the first 9 ps of the reflected signal cannot be used to infer strong coupling; coupling signatures appear only after the first echo returns.","The Rabi-oscillation period in the thin sample, about 9 ps, is shorter than the 22 ps period in the thick sample, consistent with smaller cavity mode volume and higher electron density.","The delay equal to one round trip means the formation time of a polariton is set by the cavity geometry, not by the matter oscillator alone.","Phase-resolved THz spectroscopy resolves the coherent energy exchange directly, so spectral splitting and time-domain beating can be compared in one experiment.","Faraday rotation reaching 90 degrees at polariton frequencies provides a polarization-based marker of the strong-coupling condition."],"supporting_citations":[{"why":"Establishes collective non-perturbative coupling of a 2DEG cyclotron resonance to high-quality THz cavity photons, the physical regime under study.","marker":"[36]"},{"why":"Demonstrates phase-resolved THz-TDS observation of phonon Rabi oscillations with a tunable cavity, the method this paper extends to Landau polaritons.","marker":"[17]"},{"why":"Accounts for the vacuum Bloch-Siegert shift and the middle zigzag mode in Landau-polariton Faraday spectra.","marker":"[37]"},{"why":"Documents superradiant decay of cyclotron resonance in 2D electron gases, the matter excitation whose time-domain response is measured.","marker":"[29]"},{"why":"Shows total Faraday rotation in a 2D electron gas, supporting the interpretation of 90-degree Faraday rotation at polariton modes.","marker":"[43]"},{"why":"Provides the general vacuum-Rabi-splitting framework for strong light-matter coupling that the paper's time-domain observation realizes.","marker":"[1]"}],"fun_headline_variants":["Landau polaritons wait a full cavity round-trip before beating","Rabi beat in Landau polaritons starts one round-trip late","THz probe shows polariton beat begins after cavity forms","Strong coupling in THz sets up only after one cavity round-trip","Cyclotron-cavity beat delayed by cavity round-trip time"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion relies on reading the first 9 ps (and the extrapolated first node near 4 ps) as genuinely coupling-free; if that similarity is just limited signal-to-noise or a phase-model artifact, the delay claim fails.","fun_headline_variants_meta":{"raw":{"variants":["Landau polaritons wait a full cavity round-trip before beating","Rabi beat in Landau polaritons starts one round-trip late","THz probe shows polariton beat begins after cavity forms","Strong coupling in THz sets up only after one cavity round-trip","Cyclotron-cavity beat delayed by cavity round-trip time"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000594,"raw_usage":{"total_tokens":2747,"prompt_tokens":874,"completion_tokens":1873,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":490,"completion_tokens_details":{"reasoning_tokens":1779}},"tokens_in":490,"tokens_out":1873,"duration_ms":13198,"temperature":1.0,"reasoning_tokens":1779,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T04:13:06.343623+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A measurement of the thick sample with improved signal-to-noise that resolves a Rabi-beating node within the first 9 ps, near 5.5 ps for a 22 ps Rabi period, would disprove the one-round-trip delay.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates phase-resolved THz-TDS observation of phonon Rabi oscillations with a tunable cavity, the method this paper extends to Landau polaritons."},{"cited_title":"Zhang, T","cited_arxiv_id":null,"evidence_quote":"Documents superradiant decay of cyclotron resonance in 2D electron gases, the matter excitation whose time-domain response is measured."},{"cited_title":"Khitrova, H","cited_arxiv_id":null,"evidence_quote":"Provides the general vacuum-Rabi-splitting framework for strong light-matter coupling that the paper's time-domain observation realizes."}],"review_version":2}