{"id":"71133eda-0209-4653-8fbb-cd8038f881fc","arxiv_id":"2607.14565","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A voltage-controlled Bragg mirror in a thin-film lithium niobate waveguide switches reflectivity from near zero to near unity and modulates at about one gigahertz.","lead":"Researchers built a tiny on-chip mirror whose reflectivity can be turned from nearly zero to nearly total with a voltage, and switched at gigahertz speeds. It uses thin-film lithium niobate with alternating crystal domains to create a light-reflecting grating only when voltage is applied.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Linear EO coefficient reversal across poled domains is inferred, not directly measured; quantitative fits hinge on this untested assumption.","rationale":"The reader correctly identified that the periodic reversal of the linear EO coefficient, and its completeness, is the load-bearing assumption. The paper's structural evidence (SHG) confirms sign reversal of χ(2) but not the linear EO response, and the electrical data show unexplained offsets and reduced efficiency (α). These gaps justify a CONDITIONAL verdict. My stress test agrees with this assessment and proposes a direct PFM measurement as the concrete check that would settle the concern. No additional fatal flaw was found; the central claim (voltage-controlled, GHz-speed DBR reflection) is plausible and supported by coherent measurements, but the quantitative accuracy and the physical mechanism's completeness remain unverified. Hence the verdict should remain CONDITIONAL/UNCHANGED, pending this experimental validation.","tokens_in":10604,"tokens_out":6662,"duration_ms":79581,"concrete_test":"Perform piezoresponse force microscopy (PFM) on a poled waveguide (or a calibration structure poled under identical conditions) to directly map the local piezoelectric coefficient, which is linearly related to the linear EO coefficient. Verify that adjacent domains show opposite sign and equal magnitude, and extract the true duty cycle and inversion depth. Correlate this map with the measured α and V0: if the duty cycle differs from 50% or inversion is incomplete, the fitted tanh² parameters should be recalculated using the measured spatial profile, not a uniform ideal grating. This directly settles whether the voltage-induced index grating has the expected push-pull symmetry and whether the residual discrepancy with simulation is explained by poling quality.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central mechanism claims that voltage-controlled Bragg reflection arises from a periodic reversal of the linear electro-optic (EO) coefficient in poled TFLN, with ~50% duty cycle and near-complete inversion. The paper relies on SHG microscopy to infer domain quality, but SHG only demonstrates that χ(2) alternates sign (domain walls appear as dark lines); it does not measure the sign or magnitude of the linear EO coefficient r_eff that actually generates the index grating. The electrical measurements themselves reveal a residual zero-bias reflectivity and a fitted voltage offset V0 = −5.16 V, plus a measured α = 11.95 (m·V)^−1 versus 18.34 (m·V)^−1 from simulation after thickness correction—a 35% discrepancy. The authors attribute this to 'residual deviations from a 50% poling duty cycle and incomplete domain inversion,' but no quantitative domain-inversion or EO-coefficient measurement is provided. The tanh²(α(V−V0)L) model used to fit reflectivity assumes a uniform grating with a single effective α and an offset that merely shifts the voltage origin. If the poling is indeed incomplete or asymmetric, this model may still fit the data well while masking the actual physical mechanism, making the quantitative claims about efficiency and 'near-unity' reflectivity less secure. The core claim depends on this unverified assumption, and the acknowledged discrepancies show it is the weakest link.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an electro-optically programmable distributed Bragg reflector (DBR) in a periodically poled thin-film lithium niobate (TFLN) waveguide. The device applies a bias voltage across a ferroelectric domain grating; the periodic reversal of the linear electro-optic coefficient produces a voltage-controlled push-pull index grating, turning the waveguide into a DBR. Static characterization shows reflectivity tunable from near zero to near-unity, a stable center wavelength, long-term operation, and linear center-wavelength tuning with poling period. High-frequency measurements show RC-limited 3-dB roll-offs near 1 GHz in both reflection and transmission, with detuning controls confirming that the modulation originates from the dynamic DBR response.","tokens_in":10951,"tokens_out":7423,"duration_ms":84834,"significance":"If the mechanism is fully validated, this is an important advance: it provides a path to actively programmable DBRs with subwavelength index-modulation resolution set by ferroelectric domains, a capability that is difficult to achieve with thermo-optic, electro-optic, or mechanical tuning. The experimental evidence is broad: voltage-swept spectra, tanh^2 voltage scaling, wavelength stability, period-based wavelength tuning, long-term stability, and gigahertz-speed modulation with appropriate control measurements. The simulated coupling constant from geometry and bulk electro-optic coefficients is a useful design tool. However, the quantitative claims rest on the assumption that periodic poling produces complete reversal of the linear electro-optic coefficient with ~50% duty cycle; this is inferred from SHG contrast rather than directly measured, and the paper acknowledges a 35% discrepancy between the measured and simulated coupling efficiency. The core concept is plausible and the data support it, but the efficiency and mechanism claims are not yet fully secured.","major_comments":[{"comment":"The paper infers 'near-complete domain inversion' from confocal SHG microscopy. SHG probes the sign and magnitude of the second-order nonlinearity, but the DBR mechanism in Section I relies on reversal of the linear electro-optic coefficient. The two are related but can be decoupled by domain-wall space charge, incomplete inversion, or duty-cycle asymmetry. This is not a merely academic concern: the electrical data show V0=-5.16 V, residual zero-bias reflectivity, and a measured alpha about 35% below the simulated value. Please provide a direct measurement of the linear EO response per domain (e.g., spatially resolved phase/Pockels measurement) or a quantitative model of poling nonideality that reproduces these observations. Without this, the link between poling quality and the quoted tuning efficiency is not established.","section":"Section II (Fig. 2c)"},{"comment":"The fit to R=tanh^2(alpha(V-V0)L) yields alpha=11.95 (m*V)^-1 and V0=-5.16 V. The manuscript compares this with a simulated alpha=18.34 (m*V)^-1 obtained after 'accounting for film thickness nonuniformity,' but it does not report the effective grating length used, the fit residuals, or confidence intervals. A simple internal consistency check is missing: at V=0, the fitted parameters imply R=tanh^2(11.95*5.16*0.006) ~ 0.13, which should be compared with the measured zero-bias reflectivity. Please provide these details so the reader can judge whether the 35% discrepancy is physically meaningful or a consequence of how the effective length is folded into alpha.","section":"Section III (Fig. 3c)"},{"comment":"The film-thickness nonuniformity is used to explain the double-peaked transfer function and to reduce the theoretical alpha from 27.54 to 18.34 (m*V)^-1. However, the claim that the simulations 'accurately reproduce' the data is not quantified; no residual plot, error metric, or sensitivity analysis is shown. Because this correction directly changes the benchmark against which the measured alpha is compared, please include a quantitative comparison (e.g., overlay with residuals) and a sensitivity analysis of the adjusted alpha to the assumed thickness profile and to the effective-grating-length reduction.","section":"Section III (Fig. 3b)"}],"minor_comments":[{"comment":"The text refers to 'EOS 21' where the standard notation is S21; please correct.","section":"Section IV (Fig. 4b)"},{"comment":"The phrase 'from zero to near-unity' could be misread as starting at zero applied voltage. The data indicate that the reflectivity minimum occurs at the offset voltage V0 ~ -5 V, not at 0 V. Please make this explicit.","section":"Abstract / Section III"},{"comment":"The domain duty-cycle extraction from SHG is reported as 'roughly 50%' but no uncertainty or analysis of the extraction method is given. Please provide the error bars and describe the procedure in the main text or SI.","section":"Section II (Fig. 2c)"},{"comment":"The gaps in the reflectivity-versus-voltage data are attributed to 'conditions of near-zero reflectivity,' but it would be helpful to state the threshold below which the reflectivity cannot be reliably identified.","section":"Section III (Fig. 3c)"}],"recommendation":"major_revision","confidential_remarks":"This is a strong and timely experimental paper with a novel concept. My main concern is the indirect inference of linear EO coefficient reversal from SHG data, and the resulting ambiguity in the quantitative efficiency comparison. The requested additional measurements and analysis should be within the scope of a revision; if they support the authors' interpretation, the paper would merit acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core result is real and new: a periodically poled TFLN waveguide whose Bragg reflectivity can be switched from near-zero to near-unity at ~1 GHz via the linear electro-optic effect. Prior DBR tuning was global index shifts or phase-change/MEMS at MHz; this is a different axis—grating strength—and it works at telecom wavelengths in a standard TFLN platform. That alone makes the paper worth reading carefully.\n\nThe static and high-frequency data are mutually consistent and rule out the obvious alternative explanations. Reflection grows monotonically with voltage, the center wavelength stays put (so it is not just index tuning), modulation appears only on the DBR features, and the 3-dB bandwidths from reflection and transmission agree at about 1 GHz. The seven-day stability is a nice practical touch. The thickness-nonuniformity story for the double-peaked spectrum is plausible and backed by metrology.\n\nThe soft spots are real but not fatal. The quantitative parameters—α ≈ 11.95 (m·V)^−1, V0 ≈ −5.16 V, the bandwidth fits—come from model fits with no error bars, no raw data, and the SI not available. The mechanism hinges on periodic poling producing a true reversal of the linear EO coefficient, but the evidence is SHG images of χ(2) domain contrast, which do not directly measure r_eff. The residual zero-bias reflection and the 35% gap between measured and simulated α are the paper's own admission that the poling is not ideal. The stress-test note is right that this is the load-bearing assumption; it is also right that the observation itself does not depend on the fit being perfect. I would want a direct measurement of r_eff in each domain, or at least a poling-duty-cycle scan with reflectivity-vs-voltage curves, before leaning on the absolute efficiency numbers. The paper already hints at a duty-cycle correlation for the zero-bias background, so the authors have the tools to close this gap.\n\nThe citation pattern looks fair: prior DBR tuning, TFLN EO, periodic poling, and relevant application references are all covered. The claims about first demonstration are appropriately scoped for what I can see; no obvious over-reach beyond the acknowledged imperfections.\n\nWho benefits: anyone working on integrated lasers, tunable filters, or programmable photonics in TFLN. It deserves a serious referee, not a desk reject. The right outcome is major revision with the missing error bars, a direct-poling-characterization experiment, and the SI. I would bring it up at reading group.","headline":"Voltage-activated DBR in poled TFLN is a genuinely new device capability with solid supporting data; the main quantitative claims rest on fits and an unverified poling-quality assumption, but the core result holds up and deserves serious referee time.","tokens_in":11483,"tokens_out":662,"would_cite":true,"duration_ms":9671,"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":"Periodically poled lithium niobate waveguides act as voltage-programmable Bragg mirrors, switching from transparent to near-total reflection at gigahertz speeds.","keywords":["electro-optic Bragg reflector","periodically poled lithium niobate","programmable photonics","distributed Bragg reflector","thin-film lithium niobate","gigahertz modulation","ferroelectric domain engineering","integrated photonics"],"falsifier":"A direct measurement of the electro-optic response in individual domains (e.g., phase-sensitive near-field microscopy under a uniform applied bias) that fails to show opposite and nearly equal index changes between adjacent domains would disprove the mechanism.","tokens_in":10511,"feed_emoji":"⚡","tokens_out":7406,"duration_ms":71215,"temperature":0.7,"pith_summary":"This paper demonstrates a new type of distributed Bragg reflector on a photonic chip whose reflectivity is set by an applied voltage rather than fixed by fabrication. The device uses a periodically poled thin-film lithium niobate waveguide: the periodic reversal of ferroelectric domains reverses the sign of the linear electro-optic coefficient, so a uniform electric field creates a periodic refractive-index grating with subwavelength resolution. Because the grating is induced electro-optically, it can switch between transparent and near-perfectly reflecting states at gigahertz speeds, and the operating wavelength is set by the poling period. The paper reports voltage-controlled reflectivity from zero to near-unity, a stable center wavelength, and roughly 1 GHz modulation bandwidth, which matters for reconfigurable photonic circuits, tunable lasers, and quantum optical devices.","feed_headline":"Voltage alone tunes a chip's reflection from zero to near-unity","feed_subtitle":"An electro-optic grating in poled lithium niobate switches the mirror state at gigahertz speeds.","key_machinery":"The central object is the electro-optic nonlinear ferroelectric grating: a periodically poled thin-film lithium niobate waveguide where alternating domains have opposite signs of χ(2) and therefore opposite linear electro-optic coefficients. A bias voltage across the waveguide produces a push-pull periodic index modulation (period Λ = λ0/2neff) that acts as a Bragg grating, with coupling strength κ proportional to voltage. The key is that the spatial resolution of the index modulation is set by domain reversal, not by electrode geometry or thermal diffusion, enabling subwavelength-scale grating periods with no etched structural perturbation.","core_discovery":"The central discovery is that alternating ferroelectric domains in a thin-film lithium niobate waveguide form a programmable index grating under an applied bias. Because the electro-optic coefficient changes sign with domain orientation, a uniform field makes adjacent domains experience opposite index shifts, creating an index profile whose depth is proportional to voltage. The device behaves as a distributed Bragg reflector whose coupling constant κ = αV grows linearly with bias, so reflectivity follows tanh²(α(V−V0)L) and can be driven from essentially zero to unity with tens of volts. The authors demonstrate this with a 6-mm-long third-order poled waveguide, observing near-unity reflectiv","pith_inferences":["If the linear electro-optic coefficient is indeed periodically reversed with high fidelity, the same device concept could be extended to other ferroelectric thin films (e.g., barium titanate) to push operating speeds toward tens of gigahertz, limited mainly by electrode RC constants.","The ability to set the index modulation profile purely by domain engineering implies that aperiodic or chirped domain patterns could create arbitrarily programmable spectral filters or photonic-crystal-like structures, a generalization the paper hints at but does not develop.","The observed zero-bias reflectivity and offset voltage suggest the device can serve as a sensitive diagnostic of ferroelectric domain fidelity—the linear EO response maps directly to domain structure, so residual reflections measure duty-cycle and inversion quality.","The demonstrated 1 GHz bandwidth is RC-limited, not fundamental; traveling-wave electrode designs could extend modulation speeds far beyond this, making these gratings competitive with conventional electro-optic modulators."],"forward_implications":["Voltage-controlled Bragg reflection with zero-to-near-unity range enables electrically reconfigurable mirrors for integrated lasers, allowing in-situ output power optimization and Q-switching.","Gigahertz-speed reflectivity modulation makes these gratings useful as high-speed modulators, optical switches, and tunable filters in telecom-band photonic circuits.","Since the grating period is set by poling, the center wavelength is programmable by design, as demonstrated by devices spanning roughly 1526–1608 nm from different periods.","The approach can extend to first-order poled devices with periods near 200 nm, promising higher efficiency and shorter wavelengths (including visible) when integrated with emitters.","Combining electro-optic modulation with the periodic χ(2) spatial profile opens space-time modulation physics, such as magnetic-free nonreciprocity and optical field amplification."],"fun_headline_variants":["Voltage writes a mirror on a chip","Chip mirror flips at gigahertz speed","Poled lithium niobate: reprogrammable reflector","Electro-optic grating switches light at gigabit rates","On-chip Bragg mirror tuned by a knob"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole scheme assumes that periodic poling actually reverses the linear electro-optic coefficient—not just the second-order nonlinearity—through the full depth of the waveguide, with a roughly 50% duty cycle, so a uniform applied field creates equal and opposite index shifts in neighboring domains.","fun_headline_variants_meta":{"raw":{"variants":["Voltage writes a mirror on a chip","Chip mirror flips at gigahertz speed","Poled lithium niobate: reprogrammable reflector","Electro-optic grating switches light at gigabit rates","On-chip Bragg mirror tuned by a knob"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000538,"raw_usage":{"total_tokens":2386,"prompt_tokens":681,"completion_tokens":1705,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":425,"completion_tokens_details":{"reasoning_tokens":1632}},"tokens_in":425,"tokens_out":1705,"duration_ms":12150,"temperature":1.0,"reasoning_tokens":1632,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T01:43:19.338596+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of the electro-optic response in individual domains (e.g., phase-sensitive near-field microscopy under a uniform applied bias) that fails to show opposite and nearly equal index changes between adjacent domains would disprove the mechanism.","supporting_citations":[],"review_version":1}