{"id":"34814fee-05ea-4fc8-9b5e-bc7886d9de81","arxiv_id":"2412.13830","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"By switching between multi-cycle and single-cycle THz pulses, the authors selectively excite the pi-band Higgs mode and the Leggett mode in MgB2, resolving a dispute about which mode dominates the nonlinear response.","lead":"This paper reports THz pump-probe experiments on the two-band superconductor MgB2 showing that the collective mode excited depends on the pulse shape: multi-cycle narrowband pulses select the pi-band Higgs mode, while single-cycle pulses select the Leggett mode. The result offers a protocol to identify and selectively control collective modes in multiband superconductors, potentially resolving a long-standing experimental controversy.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'selective excitation' claim is one-sided: the single-cycle experiment only shows a 1.8 THz overdamped response, but no control with a pump lacking spectral overlap with the Leggett mode is reported, so the inferred suppression of the π-band Higgs free oscillation is not demonstrated.","rationale":"The reader's weakest-assumption analysis focused on the identification of the 1.8 THz oscillation as the Leggett mode, citing the uncertain σ-gap ratio and the broad spectral peak. That concern is valid and contributes to the conditional verdict. However, the more load-bearing issue is structural: even if the frequency match were perfect, the 'selective excitation' claim requires demonstrating that the π Higgs free oscillation appears when the single-cycle pump does not overlap the Leggett mode. The paper explicitly says such a control is 'highly desirable' but does not perform it. Without that control, the absence of the π Higgs oscillation in the present single-cycle data could be a sample or measurement artifact rather than a consequence of Leggett-mode dominance. Therefore the reader's verdict of CONDITIONAL remains appropriate, but for a slightly broader reason than the one emphasized. The paper has independent strengths: the NbN control shows the setup can detect a Higgs free oscillation in a single-band superconductor; the fluence dependence and the comparison with E_pump^2 rule out trivial pump artifacts; the multicycle 2D spectroscopy identifies specific nonlinear channels; and the screening corrections are documented. These support the multicycle π-Higgs assignments more strongly than the single-cycle Leggett assignment. A missing control experiment is not grounds for rejection, but it should be a stated condition for full acceptance of the selective-excitation narrative.","tokens_in":17592,"tokens_out":5041,"duration_ms":52297,"concrete_test":"Repeat the single-cycle pump-probe measurement on the same MgB2 film with the pump spectrum shifted below roughly 0.5 THz (or with a spectral notch around 1.8 THz) while keeping τpΔπ < 1. If a free oscillation near 0.88 THz (2Δπ) then appears, the selective-excitation scenario is supported; if it does not, the absence of the π Higgs free oscillation cannot be attributed to Leggett-mode dominance, and the central claim would need to be weakened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of selective excitation requires showing that the π-band Higgs free oscillation appears when the single-cycle pump does not overlap the Leggett mode, and that it is absent when it does. The paper provides only the latter half. In the section 'Leggett mode response within non-adiabatic excitation', Fig. 3B shows a single-cycle pump whose spectrum peaks near 0.8 THz, with ωL estimated at about 1.8 THz. The text asserts 'sufficient spectral overlap' with the Leggett mode, but gives no quantitative overlap estimate, and the non-adiabatic criterion τpωL/2 ≈ 0.88 is marginal. The observed FFT feature is broad (1.8 ± 0.8 THz), and the absence of a free oscillation near 2Δπ ≈ 0.88 THz is used as supporting evidence. That absence could instead reflect that the π Higgs mode is overdamped or unresolved in this 10 nm film, independent of Leggett-mode excitation. The manuscript itself acknowledges that a single-cycle pulse without sufficient overlap with the Leggett mode would be 'highly desirable' to observe the π Higgs free oscillation, but no such control is reported. Thus the 'selective excitation' conclusion rests on a missing control condition, and the assignment of the 1.8 THz overdamped response to the Leggett mode as the dominant collective mode is not uniquely established. The Leggett-mode frequency assignment also depends on the assumed Δσ ≈ 3Δπ and literature pairing potentials, but the missing control is the more load-bearing gap.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports THz pump–probe experiments on a 10 nm MgB2 film and uses two pulse protocols to address the controversy over which collective mode dominates the nonlinear response. Under multi-cycle narrowband driving, the authors resolve nonlinear signals at the pump frequency ω and at 2ω, with temperature-dependent spectral weight showing resonances at 2ωp = 2Δπ(T) and ωp = 2Δπ(T); they attribute these to the π-band Higgs mode, supported by a two-dimensional THz spectroscopy analysis that separates the contributing nonlinear processes. Under single-cycle non-adiabatic excitation, they observe an overdamped oscillation at 1.8 ± 0.8 THz, which they assign to the Leggett mode on the basis of a computed eigenfrequency ωL ≈ 1.81 ± 0.27 THz. The central claim is that the Higgs and Leggett modes can be selectively excited by tuning the spectrum and duration of the THz pump.","tokens_in":17970,"tokens_out":5923,"duration_ms":54884,"significance":"If the mode assignments hold, the paper would resolve a genuine conflict between previous MgB2 experiments: Giorgianni et al. reported Leggett-mode dominance under single-cycle driving, while Kovalev et al. reported π-band Higgs dominance under narrowband driving. The present work attempts a unified picture within a single sample. The experimental effort is substantial: the authors carefully calibrate screening effects through measured transmission coefficients, check fluence dependence to stay in the perturbative regime, and include an NbN comparison as a positive control for the observation of a Higgs-mode free oscillation. The 2D THz spectroscopy analysis is a strength because it assigns the ω and 2ω features to specific nonlinear kernels. The paper is also explicit about a key limitation: the Summary states that observing the π-band Higgs free oscillation with a single-cycle pulse lacking Leggett-mode overlap is 'highly desirable', thereby conceding that a necessary control is missing. This transparency is commendable, but the missing control is load-bearing for the selective-excitation claim.","major_comments":[{"comment":"The assertion that the single-cycle pump has 'sufficient spectral overlap' with the Leggett mode is not quantified. The pump spectrum shown in Fig. 3B peaks near 0.8 THz and the text notes its tail extends to higher frequencies, but no measure is given of the spectral weight at ωL ≈ 1.8 THz relative to that at 2Δπ ≈ 0.88 THz. Since τpωL/2 ≈ 0.88 is only marginally in the non-adiabatic regime, a quantitative overlap estimate is needed to support the claim that the observed overdamped response is specifically due to Leggett-mode activation rather than a weak tail effect.","section":"Leggett mode response within non-adiabatic excitation (Fig. 3B)"},{"comment":"The selective-excitation conclusion rests on an incomplete comparison. The manuscript itself states that a single-cycle pulse without sufficient overlap with the Leggett mode is 'highly desirable' to observe the π-band Higgs free oscillation, but no such control is reported. Without that control, the absence of a 0.88 THz free oscillation in Fig. 3F/G constitutes negative evidence: it could reflect overdamping of the π-Higgs mode in this 10 nm film or detection sensitivity limitations rather than dominance of the Leggett mode. Demonstrating selectivity requires showing both that the Leggett mode appears when the pump overlaps it and that the π-Higgs mode appears when the pump does not.","section":"Summary"},{"comment":"The identification of the 1.8 THz overdamped oscillation as the Leggett mode is based on a frequency match between the measured feature (1.8 ± 0.8 THz) and a calculation using Δσ ≈ 3Δπ and literature pairing potentials (ωL = 1.81 ± 0.27 THz). Given the large uncertainty in the observed frequency and the sensitivity of Eq. S11 to the assumed gap ratio and pairing potentials, this is not a mode-specific fingerprint. Although the text states that the oscillation softens with temperature, a quantitative comparison with the calculated ωL(T) shown in Fig. S14 is not provided; such a comparison would substantially strengthen the assignment.","section":"Supplementary Eq. S11 and Fig. 3G"}],"minor_comments":[{"comment":"The main text refers to 'Fig. 2B and C' for the time-domain and frequency-domain waveforms, but the figure caption labels the time-domain and frequency-domain waveforms as C and D; the χ(3) plot is B. The cross-reference should be corrected.","section":"Figure 2 caption and main text"},{"comment":"The determinant detV in Eq. S11 is not explicitly defined. Specify that it is the determinant of the 2×2 pairing-potential matrix, so that the formula is self-contained.","section":"Supplementary Eq. S11"},{"comment":"There is a typo: 'Figuree 3D' should read 'Figure 3D'.","section":"Page 10, first paragraph"},{"comment":"The table note says the parameters correspond to 'the flux and temperature conditions used for plotting Fig. 3B in the main text', but the table concerns the multi-cycle 2ω signals and appears to refer to Fig. 2B. The cross-reference should be corrected.","section":"Supplementary Table S2"},{"comment":"There is a typo: 'Legget mode' should read 'Leggett mode'.","section":"Page 9, final paragraph"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and reports substantial experimental work with careful calibrations. The central claim is plausible but not yet fully proven: the missing control for the single-cycle protocol, acknowledged by the authors, is decisive for the selective-excitation conclusion. Because this control is in principle obtainable and the multi-cycle results are solid, I recommend major revision rather than rejection. The authors should either add the missing control experiment or substantially soften the selective-excitation claim and present the single-cycle result as evidence for Leggett-mode contribution rather than dominance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, this is a serious and careful attempt to resolve the MgB2 THz controversy, and the multicycle data alone are a real contribution. Second, the 'selective excitation' punchline is one step ahead of the evidence: the single-cycle leg lacks a control pump that does not overlap the Leggett mode, so the inferred suppression of the pi-band Higgs oscillation is not demonstrated.\n\nWhat's new: the same 10 nm film is studied under both multicycle narrowband and single-cycle THz pumping. The multicycle experiments reveal a resonance in the χ(3)(ωp,ωprobe,-ωprobe) channel at ωp=2Δπ(T), which is a cleaner Higgs diagnostic than the much-debated 2ω channel. They also show no Leggett resonance in the narrowband protocol across driving frequencies, and the screening corrections, fluence checks, and NbN comparison are done carefully. The Leggett frequency in the single-cycle experiment is predicted from an independent formula with literature parameters and measured gaps, so the 1.8 THz assignment is a benchmark test, not a fit.\n\nThe soft spots. The central claim that pulse duration selectively excites one mode over the other requires showing that a single-cycle pulse without spectral overlap with the Leggett mode produces the pi-Higgs free oscillation. That control is missing, and the paper says so itself: a single-cycle pump without sufficient Leggett overlap is 'highly desirable' but never implemented. The observed absence of a ~0.88 THz oscillation could instead mean the pi-Higgs mode is overdamped or unresolved in this 10 nm film. The Leggett assignment also depends on Δσ≈3Δπ and the chosen pairing potentials; the 1.8±0.8 THz feature is broad, so it is a frequency match rather than a mode-specific fingerprint. The 2ω resonance is weak and the paper concedes the Higgs vs BCS fluctuation origin there is debated; they shift weight to the ω resonance, which is reasonable but rests on a recent theoretical attribution.\n\nProportion: these are not fatal. The multicycle half of the paper is solid and the single-cycle story is plausible and consistent with prior work. What the paper has not done is prove that the Leggett mode wins because of pulse duration rather than because the pi-Higgs mode simply does not oscillate visibly in this film.\n\nThe paper is for THz spectroscopists and theorists of multiband superconductivity. It deserves a serious referee, and a good referee will push for the control experiment or a softened claim. Recommend: send to review.","headline":"Solid comparative THz study with a plausible but incomplete case for selective Higgs vs Leggett excitation; the missing single-cycle control is the main gap.","tokens_in":18521,"tokens_out":4093,"would_cite":true,"duration_ms":37479,"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":"By switching between multi-cycle and single-cycle terahertz pump pulses, this paper shows that the π-band Higgs mode and the Leggett mode in the two-band superconductor MgB2 can be selectively excited and separately identified.","keywords":["MgB2","multiband superconductivity","Higgs mode","Leggett mode","terahertz pump-probe spectroscopy","nonlinear third-order response","nonadiabatic excitation","collective mode selection"],"falsifier":"Measure the single-cycle pump-probe response in MgB2 films with independently determined σ-band gaps, or with strain or doping that changes Δσ/Δπ, and check whether the observed overdamped oscillation frequency tracks the Leggett-mode formula's predicted ωL; alternatively, suppress pump spectral weight near 1.8 THz with a notch filter while keeping τpΔπ<1 and observe whether the oscillation disappears and a π-Higgs free oscillation emerges.","tokens_in":17413,"feed_emoji":"⚡","tokens_out":7749,"duration_ms":63970,"temperature":0.7,"pith_summary":"Using terahertz pump–probe spectroscopy on thin-film MgB2, this paper claims that the collective mode dominating the nonlinear response is selected by the pump pulse's duration and spectrum. Multi-cycle narrowband pulses produce nonlinear signals at the pump frequency and its second harmonic that resonate when 2Δπ(T) matches 2ωp or ωp, identifying the π-band Higgs mode as the dominant response under periodic driving. Switching to a single-cycle pulse satisfying non-adiabatic excitation produces an overdamped ~1.8 THz oscillation, assigned by frequency matching to the Leggett mode. The two excitation protocols would thereby resolve a controversy between earlier experiments on MgB2 and give a route to distinguish Higgs modes from BCS single-particle fluctuations in multiband superconductors.","feed_headline":"Pulse shape decides between Higgs and Leggett modes in MgB2","feed_subtitle":"Multi-cycle THz light excites the π-band Higgs mode; single-cycle pulses drive the Leggett mode.","key_machinery":"The central objects are two excitation protocols: periodic driving by a narrowband multi-cycle THz pump, and non-adiabatic excitation by a single-cycle pump with τpΔ<1. The argument runs through the third-order nonlinear channels χ(3)(ωp,ωp,ωprobe), seen as a 2ω oscillation, and χ(3)(ωp,ωprobe,−ωprobe), seen as an ω oscillation, with temperature-scanned resonances matched to 2Δπ(T). The Leggett-mode assignment uses the two-band oscillator formula $ω_L^{2}$ = (Nσ+Nπ)/(NσNπ) · 4Vσπ Δσ(T)Δπ(T)/detV, fed with Δσ≈3Δπ and literature pairing potentials to get ωL=1.81±0.27 THz. The piece of the argument that distinguishes Higgs from BCS fluctuations is the 2D THz spectroscopy separation of rephasing and two-quantum processes at ω in the multi-cycle experiment.","core_discovery":"The paper establishes that the dominant THz nonlinear response of MgB2 changes identity with pump waveform. Under narrowband periodic driving, the χ(3)(ωp,ωp,ωprobe) and χ(3)(ωp,ωprobe,−ωprobe) signals resonate with 2Δπ(T), not with the σ-band gap or the Leggett frequency, so the π-band Higgs mode is the main contributor; the ω channel in particular is argued to select the amplitude (Higgs) mode over BCS charge fluctuations. Under single-cycle, non-adiabatic driving with τpΔπ≈0.44 and τpωL/2≈0.88, the response is an overdamped 1.8±0.8 THz oscillation that matches the computed Leggett eigenfrequency ωL=1.81±0.27 THz, and the expected π-Higgs free oscillation is absent. The conclusion is that interband coupling makes the Leggett mode the dominant non-adiabatic response, and that pulse spectrum and duration can be tuned to excite either mode.","pith_inferences":["If waveform-selective excitation is generic, the same pulse-shaping logic could be applied to iron-pnictide and nickelate multiband superconductors, where Higgs and Leggett assignments are also contested.","The paper's Leggett identification is essentially an eigenfrequency match; a cleaner test would vary the σ/π gap ratio through strain, doping, or different films and check that the observed overdamped frequency tracks the predicted ωL rather than staying fixed.","The authors' proposed two-pulse experiment—one non-adiabatic pulse to excite the Leggett mode and one periodic driving to probe the π Higgs—could directly measure Higgs–Leggett coupling, and is a natural next step they only outline.","The ω (rephasing/two-quantum) channel may serve as a generic Higgs-selective diagnostic beyond MgB2, since it is argued to suppress BCS-fluctuation backgrounds."],"forward_implications":["Under periodic driving, resonances appear only when 2ωp or ωp crosses 2Δπ(T), not at σ-gap or Leggett conditions, so the multicycle protocol isolates the π-band Higgs response.","The ω-channel resonance at ωp=2Δπ(T) offers a practical discriminator between the Higgs amplitude mode and BCS single-particle fluctuations.","Under non-adiabatic single-cycle excitation with enough spectral weight near ωL, the Leggett mode dominates and the π-Higgs free oscillation is overdamped or invisible.","The Leggett mode sets an upper frequency boundary for the pump spectrum if one wants to observe π-Higgs free oscillations in a two-band superconductor.","A single-cycle pump shaped to avoid spectral overlap with the Leggett mode should reveal the expected π-Higgs free oscillation."],"supporting_citations":[{"why":"Defines the non-adiabatic Higgs-oscillation protocol in NbN and provides the comparison material for MgB2.","marker":"[8]"},{"why":"Introduces the relative-phase Leggett mode in two-band superconductors.","marker":"[20]"},{"why":"Reports Raman observation of the Leggett mode in MgB2 and supplies the pairing potentials and densities of states used for the ωL estimate.","marker":"[24]"},{"why":"The earlier single-cycle and narrowband experiment claiming Leggett dominance that this paper re-examines and distinguishes from the multicycle Higgs response.","marker":"[25]"},{"why":"The prior multicycle THG experiment whose 2ωp=2Δπ(T) resonance the present 2ω data reproduce and extend.","marker":"[26]"},{"why":"Theoretical dirty-limit result that disorder enhances the π-band Higgs contribution, used to support the periodic-driving assignment.","marker":"[27]"},{"why":"Theoretical counterpoint attributing THG to BCS charge fluctuations and Leggett effects, motivating the use of the ω channel.","marker":"[28]"},{"why":"Terahertz 2D coherent spectroscopy work used to assign the ω channel to the amplitude (Higgs) mode.","marker":"[31]"},{"why":"Companion MgB2 2D coherent spectroscopy study supporting amplitude-mode assignment of the ω channel.","marker":"[32]"},{"why":"Theory of equivalence between non-adiabatic and periodic driving used to frame the two excitation protocols.","marker":"[33]"}],"fun_headline_variants":["Pump pulse shape selects Higgs or Leggett mode in MgB2","Multi-cycle THz hits Higgs, single-cycle hits Leggett in MgB2","Waveform control of collective modes in MgB2 superconductor","Selective Higgs and Leggett excitation by THz pulse shape","MgB2: Choose Higgs or Leggett mode with pump waveform"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the 1.8 THz overdamped oscillation is the Leggett mode because its measured frequency matches an eigenfrequency computed with Δσ≈3Δπ and literature pairing potentials, so if the actual gap ratio or interband couplings differ, the identification reduces to a frequency coincidence.","fun_headline_variants_meta":{"raw":{"variants":["Pump pulse shape selects Higgs or Leggett mode in MgB2","Multi-cycle THz hits Higgs, single-cycle hits Leggett in MgB2","Waveform control of collective modes in MgB2 superconductor","Selective Higgs and Leggett excitation by THz pulse shape","MgB2: Choose Higgs or Leggett mode with pump waveform"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000767,"raw_usage":{"total_tokens":3427,"prompt_tokens":996,"completion_tokens":2431,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":612,"completion_tokens_details":{"reasoning_tokens":2347}},"tokens_in":612,"tokens_out":2431,"duration_ms":14801,"temperature":1.0,"reasoning_tokens":2347,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:45:41.760956+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the single-cycle pump-probe response in MgB2 films with independently determined σ-band gaps, or with strain or doping that changes Δσ/Δπ, and check whether the observed overdamped oscillation frequency tracks the Leggett-mode formula's predicted ωL; alternatively, suppress pump spectral weight near 1.8 THz with a notch filter while keeping τpΔπ<1 and observe whether the oscillation disappears and a π-Higgs free oscillation emerges.","supporting_citations":[{"cited_title":"Matsunaga, Y","cited_arxiv_id":null,"evidence_quote":"Defines the non-adiabatic Higgs-oscillation protocol in NbN and provides the comparison material for MgB2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the relative-phase Leggett mode in two-band superconductors."},{"cited_title":"Blumberg, A","cited_arxiv_id":null,"evidence_quote":"Reports Raman observation of the Leggett mode in MgB2 and supplies the pairing potentials and densities of states used for the ωL estimate."},{"cited_title":"Giorgianni, T","cited_arxiv_id":null,"evidence_quote":"The earlier single-cycle and narrowband experiment claiming Leggett dominance that this paper re-examines and distinguishes from the multicycle Higgs response."},{"cited_title":"Kovalev, T","cited_arxiv_id":null,"evidence_quote":"The prior multicycle THG experiment whose 2ωp=2Δπ(T) resonance the present 2ω data reproduce and extend."},{"cited_title":"Haenel, P","cited_arxiv_id":null,"evidence_quote":"Theoretical dirty-limit result that disorder enhances the π-band Higgs contribution, used to support the periodic-driving assignment."},{"cited_title":"Fiore, M","cited_arxiv_id":null,"evidence_quote":"Theoretical counterpoint attributing THG to BCS charge fluctuations and Leggett effects, motivating the use of the ω channel."},{"cited_title":"Katsumi, J","cited_arxiv_id":null,"evidence_quote":"Terahertz 2D coherent spectroscopy work used to assign the ω channel to the amplitude (Higgs) mode."}],"review_version":1}