{"id":"1b8e91e4-d2c0-4509-8064-e54e94a749c5","arxiv_id":"2506.03359","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A lens-coupled kinetic inductance detector array with dual-polarized meandering aluminum absorbers is designed, fabricated, and tested at 6.98 THz, achieving a noise equivalent power of 8e-20 W/sqrt(Hz) and an aperture efficiency of 54%.","lead":"This paper designs, builds, and tests a new kind of detector array for far-infrared space telescopes, where each pixel is a tiny superconducting strip behind a silicon lens. The prototype reaches a noise equivalent power of 8e-20 W/sqrt(Hz), which would meet the sensitivity needs of future missions like PRIMA.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 54% aperture efficiency rests on a partial 4K filter-stack calibration; the small-aperture cross-check cannot break this degeneracy.","rationale":"The reader's weakest-assumption analysis correctly identified the absolute power calibration through the filter stack as the most load-bearing point. My independent review confirms that this is not a manufactured concern: the optical efficiency extraction in Eq. (26) depends sensitively on Fil(f), and the paper only measured the 4K transmission of one band-pass filter, relying on manufacturer data for the high-pass filters and assumed identity for the second BPF. The reconstructed 30% correction is large enough that a modest error in the HPF or second-BPF transmission would shift the inferred η_opt and hence the 54% aperture efficiency by a significant fraction. The small-aperture experiment in Appendix F does provide useful evidence that the reception pattern is modeled correctly, and I credit the paper for including it, but it cannot resolve an absolute calibration offset because it shares the same source-power calculation. I do not see a more fundamental flaw: the spectral model is grounded in prior work with full-wave validations, the fabricated geometry is characterized (e.g., line width measured by SEM), the misalignment is quantified and corrected with a separate aperture measurement, and the NEP result is plausible given previous KID work. The photon-noise-limited assumption is supported by the white noise spectra for Tbb≥18K. I therefore agree with the reader's conditional verdict; the concern warrants an explicit full-stack 4K calibration before the aperture efficiency is treated as fully confirmed, but it does not overturn the central demonstration.","tokens_in":27096,"tokens_out":5696,"duration_ms":66653,"concrete_test":"Measure the complete filter stack (both band-pass filters and both high-pass filters) at 4K in the same FTS used for the single BPF measurement, or alternatively calibrate the entire optical path end-to-end with a cryogenic blackbody source of known temperature and aperture. Re-derive η_opt for all 23 KIDs from Eq. (26) using the measured full-stack transmission instead of the reconstructed one. If the resulting η_opt values change by more than about 10%, the inferred aperture efficiency of 54% and the claimed agreement with the spectral model must be revised accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative chain from measured NEP to optical efficiency and then to the inferred 54% aperture efficiency passes through Eq. (26) in Appendix E. In that equation the numerator scales linearly with the filter-stack transmission Fil(f) while the dominant denominator term, NEP_exp,Ps^2, scales as Fil^2, so to first order η_opt ∝ 1/Fil. A systematic error in Fil therefore propagates almost inversely one-for-one into η_opt and, through η_opt = η_so·η_ap/η_f (Eq. 16), into the aperture efficiency. Section IV-B reports that only one of the two band-pass filters was measured at 4K; the two high-pass filters were modeled from manufacturer data for a 'similar filter' with a 2.5% 300K-to-4K transmission increase, and the second BPF is assumed identical to the measured one. The reconstructed 4K stack transmission raises the coupled power by 30% relative to 300K data. If the actual HPF 4K transmission differs from the assumed 2.5% increase, or if the two BPFs are not identical, the absolute power scale shifts. The small-aperture experiment in Appendix F is presented as an independent check, but it uses the same Fil(f) to compute incident power; it can validate the angular response and spillover but cannot separate a global calibration offset from the absolute efficiency scale. Because the aperture efficiency is only indirectly obtained from η_opt, this calibration uncertainty is load-bearing for the headline 54% figure. The photon-noise-limited assumption for Tbb≥18K is supported by the white noise spectra in Fig. 19 and is not the weakest link.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes an array concept in which distributed, dual-polarised meandering-strip absorbers are integrated with silicon lenses and kinetic-inductance resonators for far-infrared space astronomy. Using a previously developed spectral Floquet/plane-wave-spectrum model, the authors derive design guidelines, present lens-absorber designs for 6.98 and 12 THz, and fabricate a 25-element prototype at 6.98 THz. The experimental section reports resonator yield and quality factors, the optical coupling (normalized throughput) obtained from NEP measurements, an indirectly inferred aperture efficiency of 54% averaged over two polarizations, and a limiting NEP of 8e-20 W/sqrt(Hz) at 100 Hz modulation. The core evidence is a comparison of the measured optical efficiency per KID with model predictions, including a correction for a common lens-absorber misalignment that is tied to an independent small-aperture measurement.","tokens_in":27393,"tokens_out":8420,"duration_ms":89019,"significance":"If confirmed, the result is significant for future FIR space missions such as PRIMA: it demonstrates a lens-absorber KID focal-plane array with multi-mode optical coupling that is quantitatively understood, and it provides design rules for scaling to 12 THz. The paper's strengths are that the NEP is a genuine measured quantity, the optical-efficiency comparison is made without fitting spectral-model constants, the only ad hoc parameter is a common misalignment anchored to an independent small-aperture measurement, and the supplementary tolerance study addresses realistic fabrication uncertainties. The main weakness is that the headline 54% aperture efficiency is not measured directly but is inferred through a chain whose absolute scale depends on the calibration of the filter stack.","major_comments":[{"comment":"The absolute calibration of the filter-stack transmission Fil(f) is load-bearing for the reported aperture efficiency. In Eq. (26) the numerator is linear in Fil(f) through P_sf, while the dominant experimental denominator term NEP^2_exp,Ps scales as Fil^2; hence a fractional error in the reconstructed 4 K stack transmission propagates almost one-for-one into the inferred eta_opt and, through Eq. (16), into eta_ap. The calibration is only partial: one band-pass filter is measured at 4 K, the second is assumed identical, and the two high-pass filters are modeled from manufacturer data for a similar filter with a 2.5% correction, while the total 4 K-to-300 K power correction is 30%. Because the small-aperture experiment in Appendix F uses the same Fil(f), it can validate the angular response but cannot separate a global calibration offset from the absolute efficiency scale. Please provide a quantitative systematic uncertainty on eta_ap from this chain, or an independent absolute power calibration.","section":"Sec. IV-B and App. E, Eq. (26)"},{"comment":"The experimental validation of the 54% aperture efficiency is indirect and model-dependent. The measured quantity is eta_opt; eta_ap is obtained by dividing out the model values of the spillover and focusing efficiencies, and the agreement is conditioned on a common lens-absorber misalignment inferred from the same measurement family (Fig. 13 and Appendix F). The argument that the reception pattern is well controlled by the absorber size and lens diameter is plausible, but Appendix F checks the relative angular response using the same model quantities and the same absolute calibration. Please add an uncertainty budget for the derived eta_ap that includes the model assumptions and the sensitivity to the inferred misalignment, or demonstrate the absolute efficiency scale with a controlled illumination whose incident power is known independently.","section":"Sec. IV-B, Eq. (16), and Fig. 13"}],"minor_comments":[{"comment":"The Fig. 7 caption states that solid and dotted lines correspond to the designed and fabricated devices, while Sec. IV-A describes the fabricated-unit-cell curve as dash-dotted; please reconcile the line-style nomenclature.","section":"Fig. 7 and Sec. IV-A"},{"comment":"The phrase 'averaged over two polarisation' should read 'averaged over two polarizations'; there are a few similar grammatical slips elsewhere in the text.","section":"Abstract and Sec. I"},{"comment":"The y-axis label appears garbled as '[dB C/Hz]' in the typeset version; please check the rendering.","section":"Fig. 12(c)"},{"comment":"The NEP spectra for all KIDs are shown without any uncertainty indication; please state the measurement uncertainty or reproducibility of the NEP values in the text or add error bars.","section":"Sec. IV-B and Fig. 20"},{"comment":"The symbols g_g and g_c are used without definition; please state explicitly that these are the gap and ground-plane contributions to the kinetic inductance per unit length.","section":"Appendix C, Eqs. (8)-(10)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a good fit for IEEE Transactions on Terahertz Science and Technology. My main concern is the absolute calibration of the filter stack, which is load-bearing for the 54% aperture-efficiency claim; I recommend major revision rather than rejection because the central measurement is sound and the calibration issue is addressable within the scope of the paper. I did not see citation or novelty problems, and the model is properly attributed to earlier work by the same group."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. This paper reports the first working lens-absorber KID array, and the experimental work is the real deal. The NEP measurement, the noise spectra, and the optical coupling data are direct and credible. The 54% aperture efficiency is a different story: it's inferred from the optical efficiency, and the inference passes through a filter-stack transmission that was only partially characterized at 4 K. That number is not wrong, but it's softer than the abstract implies.\n\nWhat's genuinely new: the dual-meander absorber unit cell with dual polarization, and the full design-to-fabrication cycle at 6.98 THz. The spectral modeling technique comes from earlier work, but the application to lens-absorber KIDs, with the design guidelines and the tolerance study, is new. The fabrication is described in enough detail to be reproduced, and the measurement setup is thorough. The white noise spectra justify the photon-noise-limited assumption for Tbb above 18 K, and the small-aperture measurement is a good sanity check on the reception pattern.\n\nThe soft spots are the ones the reader flagged. The aperture efficiency is not measured directly; it's obtained from eta_opt via Eq. (16). As the stress-test note correctly points out, eta_opt scales inversely with the assumed filter transmission Fil. Only one band-pass filter was measured at 4 K; the two high-pass filters and the second band-pass are assumed or taken from manufacturer data. The correction raises the coupled power by 30%, so any residual error in that correction shifts the inferred aperture efficiency by a similar fraction. The small-aperture experiment cannot separate this global calibration offset because it uses the same Fil. Missing error bars on the NEP and efficiencies make it hard to gauge the uncertainty. These are fixable issues, not fatal ones. The central claim—that lens-absorber KIDs work and reach PRIMA-level sensitivity—holds up.\n\nI'd send this to review. An expert referee can push for a full 4 K filter characterization, or at least a systematic error budget. I would cite the NEP and the design approach, though I'd treat the 54% as provisional until the calibration is tightened.","headline":"First lens-absorber KID array works, but the 54% aperture efficiency is shakier than the abstract suggests.","tokens_in":27994,"tokens_out":5720,"would_cite":true,"duration_ms":62842,"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 fabricated 25-element array of lens-absorber KIDs at 6.98 THz matches the spectral model's predicted optical coupling, indirectly validating a 54% aperture efficiency and reaching a limiting NEP of $8\\times10^{-20}$ W/√Hz.","keywords":["kinetic inductance detectors","far-infrared astronomy","lens focal plane array","distributed absorber","dual-polarized absorber","terahertz detectors","spectral modelling","noise equivalent power"],"falsifier":"Measure the transmission of the complete filter stack (both band-pass and both high-pass filters) at 4 K with a cryogenic Fourier-transform spectrometer and recompute the absolute power incident on the detectors; if the inferred optical efficiency shifts by more than the reported scatter, the agreement with the model—and the indirectly validated $54\\%$ aperture efficiency—would not hold. A more direct check is a coherent-source measurement of the aperture efficiency at $6.98$ THz that does not depend on photon-noise-limited behavior.","tokens_in":26929,"feed_emoji":"🔭","tokens_out":7341,"duration_ms":78516,"temperature":0.7,"pith_summary":"This paper proposes a detector concept for far-infrared space telescopes: a kinetic inductance detector (a superconducting microwave resonator whose frequency shifts when radiation breaks Cooper pairs) whose inductive meander directly absorbs the incoming light, placed under a small silicon lens that concentrates radiation onto it. It argues that this lens-absorber design can reach the background-limited sensitivity ($\\mathrm{NEP} < 10^{-19}\\ \\mathrm{W/\\sqrt{Hz}}$) that actively cooled observatories require, while being easier to fabricate and assemble than antenna-coupled detectors at frequencies above 5 THz. The authors give co-design guidelines—matching the absorber's spectral absorption response to the angular range of the lens—and apply them to two designs at $6.98$ and $12$ THz. They fabricate the $6.98$ THz design as a 25-element array, measure its optical coupling to an incoherent blackbody source, and report agreement with the spectral model; this indirectly validates the predicted aperture efficiency of $54\\%$ averaged over two polarizations. They also measure a limiting noise equivalent power of $8\\times10^{-20}\\ \\mathrm{W/\\sqrt{Hz}}$ at 100 Hz modulation, meeting the stated sensitivity target.","feed_headline":"THz detector array validates 54% aperture efficiency","feed_subtitle":"Lens-coupled KIDs reach background-limited noise of 8e-20 W/√Hz, key for cooled space telescopes.","key_machinery":"The central object is the dual-polarized distributed absorber: a tightly periodic array of two meandering aluminum strips, about 380 nm wide in the fabricated device, whose combined sheet impedance is tuned so the absorber resonates against the unit cell's own capacitance and inductance at the target THz frequency. The absorber is shorted to ground on one side and joined to an open-ended NbTiN coplanar-waveguide resonator on the other, making the meander the inductive part of the KID. Around it, the design and analysis rest on a Floquet-wave equivalent circuit: the lens converts an incident plane wave into a spectrum of plane waves (a plane-wave spectrum) limited by the lens subtended angle $\\theta_L$; each spectral component is processed through the absorber's two-port impedance network; and coherent summation gives the absorbed power and the reception power pattern. The design guideline is to co-choose the lens $f$-number and the absorber spectral response so the aperture efficiency is maximized while the absorber stays small enough for dense focal-plane packing and low sensitivity to stray light.","core_discovery":"The central claim is that a KID whose aluminum meander acts simultaneously as the radiation absorber and as the resonator's kinetic inductor, sitting at the focal plane of an elliptical silicon lens, is accurately described by a spectral Floquet-wave transmission-line model, and that the fabricated prototype behaves as the model says. The measured optical coupling—how much of the power from an incoherent distributed source is absorbed, relative to a single spatial mode and polarization—matches the model for the 23 working detectors (a 92% yield). Because the reception power pattern is controlled almost entirely by the lens diameter and absorber size, the paper argues the measured optical coupling leaves no room for the aperture efficiency to differ from the model's value, which is $54\\%$ for the fabricated geometry. The limiting NEP of $8\\times10^{-20}\\ \\mathrm{W/\\sqrt{Hz}}$ at 100 Hz modulation is reached when the detector noise is dominated by photon shot noise and wave-bunching noise, i.e., the fundamental limit set by the background radiation. On its own terms, the paper establishes the lens-absorber KID as a scalable, background-limited detector for future FIR space missions, with relaxed alignment tolerances compared with lens-antenna KIDs.","pith_inferences":["A direct cryogenic measurement of both high-pass filters would be the quickest way to test the paper's power calibration; the authors' own data show the 4 K correction changes the coupled power by 30%, so the remaining unmeasured filters bound the systematic uncertainty of the 54% aperture efficiency.","The paper quotes NEP at 100 Hz modulation because a 1/f component dominates below that; whether the array delivers its sensitivity in a realistic spectral survey depends on how that 1/f noise is handled at the slow chop frequencies typical of space observations.","Since the two meander orientations are quasi-independently tunable, a natural extension—not pursued here—is a polarization-sensitive variant that reads out the two orientations separately.","The plane-wave illumination used in the validation is a single-direction idealization; a full end-to-end test with a real reflector feed would show whether the reception power pattern, including the measured 4.6° beam tilt from misalignment, delivers the predicted array observing speed."],"forward_implications":["The same spectral-model workflow can be reused for other bands: the 12 THz design extends the dual-meander absorber family to the 25–30 µm range with predicted 1-dB bandwidth of 5.1 THz.","The reported 92% array yield and agreement of measured versus modeled optical coupling support scaling the approach to kilo-pixel arrays read out by a single microwave backend.","At 100 Hz modulation the limiting NEP is $8\\times10^{-20}\\ \\mathrm{W/\\sqrt{Hz}}$, below the $10^{-19}$ threshold needed for background-limited spectroscopy from an actively cooled telescope.","Because the tolerated lateral misalignment for lens absorbers is far looser than for lens antennas, the design relaxes the assembly precision required for FIR focal planes.","The measured optical efficiency of about 5.95 times a single-mode, single-polarization coupling is consistent with the multi-mode effective-area model, meaning the same formulas can predict how a full instrument will couple to an extended astronomical source."],"supporting_citations":[{"why":"Supplies the spectral-model foundation: representing the focusing optics' focal field as a plane-wave spectrum.","marker":"[22]"},{"why":"Derives the figures of merit (aperture efficiency, reception power pattern, focusing efficiency) that the paper uses and validates them against full-wave simulations.","marker":"[19]"},{"why":"Earlier experimental validation of the spectral technique on a dual-band absorber bolometer array, establishing the model's credibility.","marker":"[23]"},{"why":"Provides the measurement methodology for NEP and optical efficiency from photon-noise-limited KIDs, including the filter-stack correction approach.","marker":"[13]"},{"why":"The nearest comparable detector (a 12 THz single-photon-sensitive KID) used as performance benchmark and comparison for NEP and quasiparticle lifetime.","marker":"[14]"},{"why":"Defines normalized throughput for filled and feedhorn-coupled focal planes, the metric the paper uses for optical coupling.","marker":"[26]"},{"why":"Establishes the criterion for photon-noise-limited operation (white noise with quasiparticle-lifetime roll-off) that justifies extracting optical efficiency from NEP.","marker":"[41]"},{"why":"Gives the cascade formulation the paper cites for extending the lens-absorber model to a full reflector-based imaging system.","marker":"[24]"}],"fun_headline_variants":["Lens-coupled KIDs hit 54% aperture efficiency","THz KIDs achieve background-limited noise for space","Dual-polarized KIDs validate 54% aperture efficiency in THz","Scalable KID array passes 54% efficiency test for far-IR space","Lens absorber KIDs reach shot-noise limit at 8e-20 W/√Hz"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The measured optical efficiency, and thus the indirectly inferred aperture efficiency of $54\\%$, rests on the assumptions that the KIDs are photon-noise limited for blackbody temperatures above 18 K and that the absolute power incident on the detectors is known; the filter-stack transmission was calibrated with a 4 K measurement of only one band-pass filter, while the two high-pass filters were modeled from manufacturer data.","fun_headline_variants_meta":{"raw":{"variants":["Lens-coupled KIDs hit 54% aperture efficiency","THz KIDs achieve background-limited noise for space","Dual-polarized KIDs validate 54% aperture efficiency in THz","Scalable KID array passes 54% efficiency test for far-IR space","Lens absorber KIDs reach shot-noise limit at 8e-20 W/√Hz"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000246,"raw_usage":{"total_tokens":1571,"prompt_tokens":1005,"completion_tokens":566,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":621,"completion_tokens_details":{"reasoning_tokens":470}},"tokens_in":621,"tokens_out":566,"duration_ms":6162,"temperature":1.0,"reasoning_tokens":470,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:04:28.269136+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the transmission of the complete filter stack (both band-pass and both high-pass filters) at 4 K with a cryogenic Fourier-transform spectrometer and recompute the absolute power incident on the detectors; if the inferred optical efficiency shifts by more than the reported scatter, the agreement with the model—and the indirectly validated $54\\%$ aperture efficiency—would not hold. A more direct check is a coherent-source measurement of the aperture efficiency at $6.98$ THz that does not depend on photon-noise-limited behavior.","supporting_citations":[{"cited_title":"Llombart, B","cited_arxiv_id":null,"evidence_quote":"Supplies the spectral-model foundation: representing the focusing optics' focal field as a plane-wave spectrum."},{"cited_title":"Llombart, S","cited_arxiv_id":null,"evidence_quote":"Derives the figures of merit (aperture efficiency, reception power pattern, focusing efficiency) that the paper uses and validates them against full-wave simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the measurement methodology for NEP and optical efficiency from photon-noise-limited KIDs, including the filter-stack correction approach."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The nearest comparable detector (a 12 THz single-photon-sensitive KID) used as performance benchmark and comparison for NEP and quasiparticle lifetime."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines normalized throughput for filled and feedhorn-coupled focal planes, the metric the paper uses for optical coupling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the cascade formulation the paper cites for extending the lens-absorber model to a full reflector-based imaging system."}],"review_version":1}