{"id":"88a8e28b-d336-480e-ab5b-8a37650d5904","arxiv_id":"1908.09581","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Full 3D+1 simulations show that in tilted-pulse-front terahertz sources, few-cycle pulses are generated only near the prism apex, with beam-size-dependent spatial inhomogeneity.","lead":"This paper models, in full 3D, how terahertz pulses are generated by a tilted-pulse-front setup and finds the output is spatially uneven: clean few-cycle pulses form only near the sharp edge of the crystal. The result matters because strong-field terahertz experiments need well-characterized pulses, and it gives beam-size rules for making them.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Quantitative claims lack numerical convergence or validation; a resolution test is needed before the 25% and energy-fraction results can be trusted.","rationale":"I read this paper as a simulation-based claim: solving Eqs. (5)-(6) in 3D+1 reveals spatial inhomogeneity of the THz field and shows that 1D+1/2D+1 models overestimate conversion efficiency. For this claim to hold, the numerical solver must be demonstrated to resolve the relevant spatial and spectral scales. The paper does not provide that demonstration: there is no convergence study, no stated grid resolution, no time-step or spectral-window convergence, and no code or data release. The reader identified SVEA as the weakest assumption, but in the frequency-domain formulation used here the validity condition is spatial slowness along z, not temporal envelope slowness, and the large THz absorption lengths make the neglect of ∂²/∂z² plausible for the reported frequencies. The actual gap in the evidence chain is numerical verification: the quantitative numbers in the abstract and figures are presented with no error control. This supports the reader's CONDITIONAL verdict rather than changing it, because a convergence test could be added without altering the qualitative conclusions.","tokens_in":10494,"tokens_out":21897,"duration_ms":242930,"concrete_test":"Rerun the 3D+1 simulation for one of the three beam sizes (e.g. σx'=0.88 mm) with at least 2× higher transverse resolution in x and y and 2× smaller longitudinal step Δz, and compare the total conversion efficiency, the apex THz temporal waveform, and the percentage of energy with ∆t>2∆t_ref. Also rerun with a wider spectral window; if any of these change by more than a few percent, the reported numbers are not converged and the central claims need qualification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative conclusions—the ~25% efficiency gap between 2D+1 and 3D+1 and the 4%/20%/25% non-single-cycle energy fractions—rest entirely on the accuracy of the FFT-BPM/split-step solution of Eqs. (5)-(6). The paper reports no convergence checks, no grid or step-size settings, no spectral-window tests, and no comparison to an independent solver or experiment. The claim that y-diffraction is negligible for σy∈[0.5,4.5] mm is asserted '(not shown)'. If the transverse grid is too coarse or the longitudinal steps too large, the spatial sharpening near the prism apex and the grey-region fractions in Fig. 6 could be numerical artifacts rather than physical. This is more load-bearing than the reader's SVEA concern: the frequency-domain equations in (5)-(6) are a unidirectional propagation form, and the SVEA-in-z condition is plausibly satisfied for the THz absorption lengths here; the missing error control is the part of the evidence chain that is actually undocumented.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a 3D+1 numerical model of tilted-pulse-front terahertz generation in lithium niobate, based on coupled-wave equations solved with FFT-BPM and split-step Fourier methods. It first derives analytically that the grating's second-order angular dispersion produces a spatial narrowing of the generated terahertz bandwidth (Eq. 4), then compares 1D+1, 2D+1, and 3D+1 simulations. The central claims are that the pump beam size in the pulse-front-tilt plane strongly affects the spatio-temporal THz field, that few-cycle pulses occur only near the prism apex, that the 1D+1 and 2D+1 models overestimate conversion efficiency, with 2D+1 overestimating by about 25% relative to 3D+1, and that the non-single-cycle region contains 4%, 20%, and 25% of the THz energy for the three studied pump sizes. No code or data are shipped, and no numerical convergence or validation tests are reported.","tokens_in":10711,"tokens_out":2430,"duration_ms":25832,"significance":"If the quantitative results hold, the paper makes a useful contribution to the design of tilted-pulse-front THz sources: it identifies a practical limitation of reduced-dimensional models and quantifies the spatial inhomogeneity that matters for strong-field and carrier-envelope-phase-sensitive applications. The analytic derivation in Section 2 is clean and dimensionally consistent, and the numerical model includes relevant physics: χ(2) generation and back-conversion, cascading, n2, self-steepening, and stimulated Raman effects. The comparison of 1D+1, 2D+1, and 3D+1 models is timely and likely of interest to the community. However, the paper's quantitative conclusions rest entirely on an undocumented numerical implementation, which limits confidence until convergence and validation evidence is provided.","major_comments":[{"comment":"The central quantitative claims—the ~25% efficiency gap between 2D+1 and 3D+1, and the 4%/20%/25% non-single-cycle energy fractions in Fig. 6—depend on the numerical accuracy of the FFT-BPM/split-step solution, but the manuscript reports no grid sizes, step sizes, spectral windows, absorption-boundary settings, or convergence tests. Without a resolution study or a comparison to an independent solver or experimental data, it is unclear whether the sharp spatial features near the prism apex and the grey-region energy fractions are physical or numerical artifacts. A convergence analysis, at least for the key reported quantities, is required before these numbers can be accepted.","section":"Section 3, Eqs. (5)-(6)"},{"comment":"The claim that y-diffraction has negligible effect on THz generation for σy in [0.5, 4.5] mm is stated as '(not shown)'. This assertion is load-bearing because it justifies the use of the 2D+1 model for the beam-size scan in Fig. 3 and the conclusion that a 2D calculation is a good approximation. The supporting data or an analytic estimate should be shown, or the claim should be softened.","section":"Section 4, first paragraph"},{"comment":"The coupled-wave model applies the slowly varying amplitude approximation to the THz field. For single- to few-cycle THz pulses the envelope changes on the same timescale as the carrier, so SVEA may be inaccurate for the broadband THz field. The paper does not discuss or test this approximation, even though the quantitative spatial profiles depend on it. A concrete test would be to compare with a unidirectional pulse propagation formulation without the SVEA, or to check the validity condition |∂zE| << k0|E| over the THz spectrum for the simulated parameters.","section":"Eqs. (5)-(6) and Fig. 5"}],"minor_comments":[{"comment":"Equation (7) contains a typo in the reference ('see Eq.7)' missing a space) and the notation t(x,y)p should be defined more clearly, since t is used both as the time coordinate and in the subscript p.","section":"Eq. (7)"},{"comment":"The caption says 'calculated by the 2D model' while the text uses '2D+1 model' consistently elsewhere; the caption should match the text.","section":"Fig. 3 caption"},{"comment":"The parameter 'focal length f2' is given as '0.613× f1 mm [15]'; the formatting is confusing and the unit mm appears to apply to f1 only. Please clarify the value and the reference.","section":"Table 1"},{"comment":"The phrase 'wasted' in the discussion of large OP beam sizes is fine colloquially, but the preceding sentence about absorption would be clearer if it stated that the THz generated near the base is reabsorbed before reaching the output surface, which is already mentioned and needs no change; this is a wording suggestion only.","section":"Section 4, Fig. 4"}],"recommendation":"major_revision","confidential_remarks":"The paper's main scientific message is plausible and relevant, and the analytic part is solid. The missing numerical convergence and validation details are, in my view, the only blocking issue; they should be added rather than requiring a full rewrite. I would also gently recommend that the authors double-check the 25% efficiency-overestimate claim against a mesh-refinement study, since this number will likely be quoted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth knowing: this is the first systematic 1D+1/2D+1/3D+1 comparison of the tilted-pulse-front setup that includes cascading and a telescope imaging system. The main finding—few-cycle THz content is confined near the prism apex, and lower-dimensional models overestimate efficiency (2D+1 by roughly 25%)—is directly useful for people designing high-field THz experiments. The analytic derivation in Eq. (4) tracing spatial bandwidth narrowing to second-order angular dispersion from the grating is clean and dimensionally consistent, and the numerics include the right physics: χ(2), cascading, n2, Raman, and frequency-dependent absorption with parameters from independent measurements.\n\nThe soft spots are real, and they are mostly about documentation. The paper ships no code, no data, no grid or step-size settings, and no convergence or spectral-window tests. The reader's worry about SVEA is secondary: the frequency-domain unidirectional equations in (5)-(6) should be fine for the absorption lengths involved. The stress-test note is the one that lands: the 4%/20%/25% energy fractions and the 25% efficiency gap are quantitative claims that rest entirely on an undocumented FFT-BPM split-step solution. Without a resolution test or a comparison to an independent solver, I'd treat those numbers as plausible but not reliable. The qualitative picture—smaller σx gives more single-cycle content, y-diffraction is weak for the tested range—will almost certainly survive a convergence check; only the exact percentages may shift.\n\nFor whom: this is for the high-field THz community, especially people pushing conversion efficiency or planning CEP-sensitive experiments. It deserves a serious referee, because the model comparison is the right question and the paper gives a clear answer. My advice: send it out, ask for a convergence study and explicit numerical parameters, and let the authors either confirm or correct the energy fractions. I'd cite it for the 3D+1-vs-2D+1 efficiency point once the numbers are pinned down.","headline":"First systematic 1D/2D/3D+1 comparison of tilted-pulse-front THz generation that is genuinely useful for experiment design, but the quantitative efficiency numbers need a convergence check before I fully trust them.","tokens_in":11282,"tokens_out":1703,"would_cite":true,"duration_ms":17004,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["42.65.Ky"],"model":"deepseek-v4-flash","headline":"Few-cycle terahertz pulses are generated only near the prism apex in tilted-pulse-front sources.","keywords":["terahertz generation","tilted pulse front","optical rectification","lithium niobate","3D+1 simulation","cascading effect","spatial inhomogeneity","conversion efficiency"],"falsifier":"Measure the spatially resolved terahertz waveform at the exit face of a lithium-niobate crystal for a pump size $\\sigma_x' = 1.32$ mm at the stated fluence; the model predicts 25% of the terahertz energy lies in the region with $\\Delta t > 2\\Delta t(x_p,y_p)$ and predicts a conversion efficiency near 0.46%. A substantially smaller non-single-cycle fraction, or an efficiency close to the 2D+1 value rather than about 25% lower, would indicate the 3D+1 model's central spatial predictions are wrong.","tokens_in":10286,"feed_emoji":"⚡","tokens_out":10788,"duration_ms":94064,"temperature":0.7,"pith_summary":"This paper establishes, through a full three-dimensional-plus-time simulation, that the spatio-temporal quality of terahertz pulses from tilted-pulse-front setups depends strongly on where in the pump beam the conversion happens. Clean few-cycle terahertz waveforms are produced only near the apex of the lithium-niobate prism; the rest of the beam, which can hold a large fraction of the energy, arrives as longer, chirped, multi-cycle fields. The paper also shows that reduced-dimensional models mislead: 1D+1 and 2D+1 both overestimate the optical-to-terahertz conversion efficiency, with 2D+1 overestimating by about 25% relative to 3D+1. This matters because strong-field, carrier-envelope-phase-sensitive terahertz experiments need to know the actual spatial composition of the pulse, not just its total energy.","feed_headline":"Terahertz few-cycle pulses form only near the crystal apex","feed_subtitle":"Larger beams in the tilt plane add multi-cycle energy; 2D models overestimate conversion efficiency by 25%.","key_machinery":"The machinery is a 3D+1 coupled-wave model that evolves the optical pump and terahertz fields in $(x,y,z,t)$ using the fast-Fourier-transform beam propagation method with split-step integration. It retains diffraction in both transverse directions, the spatial walk-off term $2ik_{x0}(\\omega)\\,\\partial/\\partial x$ that lower-dimensional models drop, cascading back-conversion between terahertz and pump, third-order nonlinearities such as self-phase modulation and stimulated Raman scattering, and frequency-dependent terahertz absorption. A second load-bearing piece is the analytic expansion of the grating angular dispersion to second order, whose coefficient $F_2$ enters the nonlinear polarization through a term proportional to $x'^2 (F_2/F_1)^2$ and produces a position-dependent terahertz bandwidth even before nonlinear propagation.","core_discovery":"The central discovery is that the terahertz waveform generated in a grating-based tilted-pulse-front setup is spatially inhomogeneous in a specific, predictable way: the few-cycle character is confined to the vicinity of the crystal apex, and the fraction of energy outside this single-cycle region grows with the pump beam size in the pulse-front-tilt plane. The paper derives analytically that second-order angular dispersion from the grating alone makes the terahertz bandwidth decrease away from the pump center, and it shows numerically that the full 3D+1 model predicts conversion efficiency about 25% lower than the 2D+1 model because only the full model accounts for pump depletion and fluence reduction along the third dimension.","pith_inferences":["The same spatial-inhomogeneity mechanism should appear in other prism-shaped nonlinear crystals and at other pump wavelengths, so the qualitative recommendation to keep the tilt-plane beam small is likely transferable beyond lithium niobate.","The analytic $F_2$ term predicts a specific spatial chirp of the terahertz spectrum across the output face; a spatially resolved measurement of terahertz spectra could be fitted to that $x'$ dependence as a direct test of the model independent of efficiency measurements.","The predicted 4%, 20%, and 25% non-single-cycle energy fractions could be checked by imaging the terahertz beam onto a segmented detector and comparing the pulse duration map; a large discrepancy would point to the slowly-varying-envelope approximation as the limiting assumption.","The paper's recommendation implies elliptical pump beams as an optimization strategy: small in the tilt plane for waveform quality, large in the perpendicular direction for energy scaling."],"forward_implications":["The 2D+1 model is adequate for predicting optical and terahertz spectra, but any quantitative conversion-efficiency claim must come from a 3D+1 calculation; 2D+1 overestimates efficiency by about 25%.","For the simulated parameters, the non-single-cycle region, defined as $\\Delta t(x,y) > 2\\Delta t(x_p,y_p)$, contains 4%, 20%, and 25% of the terahertz energy for pump sizes $\\sigma_x' = 0.44$, 0.88, and 1.32 mm.","The terahertz beam size in the direction perpendicular to the pulse-front-tilt plane scales as $\\sigma_y/\\sqrt{2}$ and is insensitive to diffraction for $\\sigma_y$ between 0.5 and 4.5 mm, so the 2D+1 approximation is good for spectral studies in that regime.","Keeping the pump small in the pulse-front-tilt plane while enlarging it in the perpendicular direction preserves single-cycle content while still scaling up total energy.","Carrier-envelope-phase-sensitive terahertz experiments should not treat the generated pulse as a single uniform few-cycle field; the energy generated away from the apex is temporally chirped and multi-cycle."],"supporting_citations":[{"why":"Prior 2D+1 dynamic theory for low-absorbing crystals that lacks back conversion; the baseline this paper says overestimates efficiency.","marker":"[13]"},{"why":"Prior 2D+1 model that includes back conversion and one-lens imaging, which the paper extends to 3D+1 with a telescope system.","marker":"[14]"},{"why":"Telescope imaging setup reference that fixes the focal-length ratio and the imaging-error trade-off used in the simulations.","marker":"[15]"},{"why":"Supplies the fast-Fourier-transform beam propagation method on which the 3D+1 numerical solver is built.","marker":"[16]"},{"why":"Provides the imaging condition that places the grating image parallel to the pulse-front-tilt plane, defining the coordinate frame of the model.","marker":"[17]"},{"why":"Analysis of tilted-pulse-front terahertz generation that supplies the second-order angular-dispersion treatment used in the analytic argument.","marker":"[20]"},{"why":"Source of the peak-fluence damage-threshold estimate that sets the optical pump fluence in the simulations.","marker":"[23]"},{"why":"Experimental result that higher pump fluence produces smaller terahertz beams, used to validate the simulated fluence trend.","marker":"[26]"},{"why":"Proposal of a stair-step echelon setup cited as the route to spatially homogeneous terahertz pulses.","marker":"[27]"}],"fun_headline_variants":["Few-cycle THz only near apex in tilted-pulse-front","3D+1 modelling: THz single-cycle only near prism apex","Larger beams in tilt plane ruin THz few-cycle character","3D+1 model: THz few-cycle only at prism apex","Beam size in tilt plane dictates THz waveform shape"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model assumes the terahertz field's envelope changes slowly compared with its carrier oscillation, even though the terahertz pulse is only one-to-few cycles; if that slowly-varying-envelope approximation is inaccurate for the broadband terahertz field, the predicted spatial fractions and efficiency numbers would shift.","fun_headline_variants_meta":{"raw":{"variants":["Few-cycle THz only near apex in tilted-pulse-front","3D+1 modelling: THz single-cycle only near prism apex","Larger beams in tilt plane ruin THz few-cycle character","3D+1 model: THz few-cycle only at prism apex","Beam size in tilt plane dictates THz waveform shape"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000697,"raw_usage":{"total_tokens":3148,"prompt_tokens":942,"completion_tokens":2206,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":558,"completion_tokens_details":{"reasoning_tokens":2117}},"tokens_in":558,"tokens_out":2206,"duration_ms":14938,"temperature":1.0,"reasoning_tokens":2117,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:06:43.722416+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the spatially resolved terahertz waveform at the exit face of a lithium-niobate crystal for a pump size $\\sigma_x' = 1.32$ mm at the stated fluence; the model predicts 25% of the terahertz energy lies in the region with $\\Delta t > 2\\Delta t(x_p,y_p)$ and predicts a conversion efficiency near 0.46%. A substantially smaller non-single-cycle fraction, or an efficiency close to the 2D+1 value rather than about 25% lower, would indicate the 3D+1 model's central spatial predictions are wrong.","supporting_citations":[{"cited_title":"Terahertz generation with tilted-front laser pulses: dynamic theory for low-absorbing crystals,","cited_arxiv_id":null,"evidence_quote":"Prior 2D+1 dynamic theory for low-absorbing crystals that lacks back conversion; the baseline this paper says overestimates efficiency."},{"cited_title":"Theory of terahertz generation by optical rectification using tilted-pulse-fronts,","cited_arxiv_id":null,"evidence_quote":"Prior 2D+1 model that includes back conversion and one-lens imaging, which the paper extends to 3D+1 with a telescope system."},{"cited_title":"Optimization of the tilted-pulse-front terahertz excitation setup containing telescope,","cited_arxiv_id":null,"evidence_quote":"Telescope imaging setup reference that fixes the focal-length ratio and the imaging-error trade-off used in the simulations."},{"cited_title":"Light propagation in graded-index optical fibers,","cited_arxiv_id":null,"evidence_quote":"Supplies the fast-Fourier-transform beam propagation method on which the 3D+1 numerical solver is built."},{"cited_title":"Designofhigh-energyterahertzsourcesbasedonopticalrectification,","cited_arxiv_id":null,"evidence_quote":"Provides the imaging condition that places the grating image parallel to the pulse-front-tilt plane, defining the coordinate frame of the model."},{"cited_title":"Analysis of terahertz generation using tilted pulse fronts,","cited_arxiv_id":null,"evidence_quote":"Analysis of tilted-pulse-front terahertz generation that supplies the second-order angular-dispersion treatment used in the analytic argument."},{"cited_title":"Pulse sequences for efficient multi-cycle terahertz generation in periodically poled lithium niobate,","cited_arxiv_id":null,"evidence_quote":"Source of the peak-fluence damage-threshold estimate that sets the optical pump fluence in the simulations."},{"cited_title":"Nonlinear distortion of intense thz beams,","cited_arxiv_id":null,"evidence_quote":"Experimental result that higher pump fluence produces smaller terahertz beams, used to validate the simulated fluence trend."},{"cited_title":"Numerical investigation of a scalable setup for efficient terahertz generation using a segmented tilted-pulse-front excitation,","cited_arxiv_id":null,"evidence_quote":"Proposal of a stair-step echelon setup cited as the route to spatially homogeneous terahertz pulses."}],"review_version":1}