{"id":"d96cd801-ffa0-4d49-8aa9-2ae96a5f6226","arxiv_id":"2607.20324","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A cryogenic, switchable photoelectric electron source made from LED and aluminum foil was demonstrated and detected at 100–300 eV with TES microcalorimeters.","lead":"A compact electron source that operates inside a cryostat was built from a commercial LED and aluminum foil, and was shown to produce 100–300 eV electrons detected by transition-edge sensor microcalorimeters. The goal is a switchable, energy-tunable calibration source for experiments like HOLMES that measure the neutrino mass.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The headline rate and efficiency claims rely on a 1D Gaussian fit to one row of a 4×16 TES array, extrapolated to a 2D profile with no uncertainty propagation; if the beam is non-Gaussian or miscentered in y, the ≥1 Hz and ≳10^-14 e-/γ numbers could be off by a large factor.","rationale":"The reader's weakest assumption identifies the same load-bearing concern I see: the quantitative rate and efficiency in the abstract rest on a beam-profile extrapolation that uses only one row of the detector array, with the other row excluded and no uncertainty propagation. This is a genuine vulnerability because the headline numbers could change by a factor of order unity or more if the beam is non-Gaussian or the y-profile differs from the x-profile. The concern is addressable with the existing dataset by summing the measured channel rates or by jointly fitting both rows with a 2D model. The paper's other evidence—polarity dependence, LED-current scaling, voltage-amplitude scaling, and energy spectra—supports the qualitative demonstration, so the appropriate verdict remains CONDITIONAL rather than REJECT. The unresolved electron escape mechanism is a notable limitation but does not directly threaten the empirical claim that electrons are being produced and detected; it is a theoretical gap rather than an observed inconsistency with the rate measurement. Therefore I agree with the reader's conditional verdict and do not recommend changing it.","tokens_in":14815,"tokens_out":12323,"duration_ms":116968,"concrete_test":"Recompute the total electron rate by summing the measured per-channel rates over all channels in the activity map (Figure 6) without a Gaussian extrapolation, after equalizing trigger thresholds for the lower row (e.g., re-fitting the noisy channel with its actual threshold). If the summed array rate is below 1 Hz, the abstract's '≥1 Hz' claim is unsupported. Alternatively, refit the upper- and lower-row rates together with a 2D Gaussian with free y-center and y-width, and compare the integrated rate with the 1D-extrapolated value; a substantially lower integrated rate would confirm that the Gaussian assumption drives the headline number.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's quantitative claims—an electron rate of ≳1 Hz and an efficiency of ≳10^-14 e−/γ—are derived in Section 4 from a beam-profile measurement that is far more fragile than the text conveys. Figure 7 shows a 1D Gaussian fit to the upper row of the TES array; the lower row was excluded because one channel had higher noise and a different trigger threshold. The fit gives a width of 1.38±0.33 mm along x, but this provides no constraint on the y-profile or the y-position of the beam center. The paper then 'assum[es] a 2D Gaussian profile' to infer the integrated source rate. If the actual beam is narrower in y, or if the beam center is not located on the upper row, the Gaussian-integrated total rate could be substantially lower than the true rate; if the beam is wider or has tails, it could be higher. No systematic uncertainty is assigned to this extrapolation. The measured per-channel rates are only tens to ~160 mHz in the most active channels, so the difference between summing the measured channels and integrating a 2D Gaussian may be the deciding factor between >1 Hz and <1 Hz. Since these numbers are explicit headline results, this is a load-bearing weakness. The qualitative demonstration of a switchable cryogenic electron source is well supported by the polarity, rate-scaling, and energy-scaling tests, but the specific rate and efficiency values are not. A secondary unresolved issue, acknowledged in Section 6, is the escape mechanism of electrons from 400-nm aluminum foils; that is an important physical puzzle but does not directly invalidate the empirical observation of electrons.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a prototype cryogenic electron source based on photoelectric emission from two 400-nm aluminum layers illuminated by a 280-nm LED, with electrons accelerated by up to 300 V and detected by a TES microcalorimeter array. The authors present three validation tests: signal appears only for negative bias, rate scales quadratically with LED current, and pulse amplitude increases with acceleration voltage. They characterize the beam profile by fitting a 1D Gaussian to the upper row of the array and, assuming a 2D Gaussian, infer a source rate of ≳1 Hz and an efficiency of ≳10^-14 e−/γ. They also analyze TES pulse energies using the Joule estimator, fit a phenomenological spectral model, and compare with Nebula simulations, reporting a measured energy resolution of about 40 eV for 100–300 eV electrons. The paper positions this as the first stage of a switchable calibration source for HOLMES and similar experiments.","tokens_in":15250,"tokens_out":5072,"duration_ms":49390,"significance":"If the quantitative claims are trustworthy, this is a valuable step toward a compact, switchable, cryogenic calibration source whose electron yield is independent of the acceleration energy—an improvement over CNT field-emission sources. The polarity, rate-scaling, and energy-scaling tests are clean and make the qualitative observation of cryogenic photoelectrons convincing. The demonstration of TES sensitivity to 100–300 eV electrons is also useful, as few such measurements exist. However, the headline rate and efficiency numbers rest on a fragile beam-profile extrapolation, and the physical mechanism of electron escape from 400-nm aluminum is explicitly acknowledged as unexplained. These issues do not undermine the qualitative demonstration but do require a more careful quantitative treatment before the specific numbers can be accepted.","major_comments":[{"comment":"The abstract's headline claims (rate ≳1 Hz, efficiency ≳10^-14 e−/γ) are derived from a 2D Gaussian integral whose width is obtained from a 1D Gaussian fit to the upper row only; the lower row was excluded. This leaves the y-profile and the beam-center y-position unconstrained. The measured per-channel rates in Fig. 6 are only 25–160 mHz, so the integrated rate depends heavily on the assumed Gaussian tail and symmetry. No systematic uncertainty is propagated from the profile assumption, and the photon-flux estimate Φ∼10^14 γ/s also lacks an uncertainty. If the beam is narrower or off-center in y, the inferred rate could be substantially lower; if it is wider or has non-Gaussian tails, higher. Please either provide a more robust rate estimate (e.g., a 2D fit using all available channels with a common threshold, or a conservative geometric-acceptance bound) or soften the abstract to qualit","section":"Section 4, Figure 7 and Abstract"},{"comment":"The paper states: 'Given the photon mean free path in aluminum and the short electron range, electrons should not be able to escape the material—contrary to our observations.' This is a serious physical gap in the central source-interpretation claim. If the standard photoemission picture cannot explain escape from 400-nm Al, then the observed electrons might originate from a different mechanism (e.g., pinholes, surface oxides, field-assisted emission, or photoemission from other surfaces). The polarity and intensity tests show that the carriers are electrons originating under negative bias, but they do not identify the emission site. Please add a discussion of possible escape mechanisms or a control measurement (e.g., varying the Al thickness or number of layers) that could distinguish the proposed interpretation. At minimum, the paper should not present the photoelectric-from-Al interpr","section":"Section 6, Conclusions"},{"comment":"The spectral model is motivated by Nebula simulations, then used to fit the data with E_J0 as a free parameter that sets the energy scale, and the same simulated curves are overlaid for comparison. This is a mild self-benchmarking loop: the simulated shape is not an independent prediction after E_J0 is fit. The 40 eV resolution is also model-dependent and the authors correctly note it is not the intrinsic TES resolution. Please clarify what E_J0 actually calibrates (e.g., whether it is tied to the known acceleration voltage) and show the sensitivity of the energy scale to the chosen functional form or to the fitted resolution. If the energy axis is not independently calibrated, this should be stated explicitly in the text.","section":"Section 5, Eqs. 2–3 and Fig. 9"}],"minor_comments":[{"comment":"The text 'Figure 2 shows a 4×16 detector array; however, the bottom two 2×16 TES arrays on the chip were not wire-bonded' is unclear. A 4×16 array is a single array; 'bottom two 2×16' suggests the chip actually contains multiple sub-arrays. Please clarify the geometry of the chip and exactly which pixels were read out.","section":"Section 2, Figure 2 caption"},{"comment":"The description of the 5σ amplitude cut should specify how the noise-event distribution was defined, whether it was per channel or global, and how the threshold was applied. This is important because the rate depends on the cut.","section":"Section 4"},{"comment":"Reference [12] (Alpert et al., Phys. Rev. Lett.) has an incomplete DOI: '10.1103/s9vl-7n24'. Please correct it.","section":"References"},{"comment":"The notation 'e−/γ' in the abstract and Section 6 is nonstandard; 'e−/photon' would be clearer. Also, the efficiency is defined relative to the nominal photon flux, but the actual flux reaching the photocathode (after fiber losses and reflection) is not quantified.","section":"General"},{"comment":"The caption states that a portion of the 100 eV spectrum has been truncated 'to maintain signal visibility'. Please specify exactly which range was truncated and how this affects the fit or the visual comparison with the simulation.","section":"Figure 9"}],"recommendation":"major_revision","confidential_remarks":"The paper's qualitative demonstration—cryogenic photoelectrons from an LED-illuminated Al photocathode, accelerated and detected by a TES array—is convincing and likely of interest to the low-temperature detector community. However, the quantitative rate/efficiency claims in the abstract are not supported by the current analysis: the 2D Gaussian extrapolation from one row has no systematic uncertainty, and the unexplained electron-escape mechanism leaves the source interpretation incomplete. The authors should either provide a robust rate estimate or explicitly downgrade the quantitative claims. The energy-scale calibration also deserves more scrutiny. I would not recommend rejection because the central observed phenomenon is well validated by the polarity and scaling tests, and the structural issues are fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a look if you care about low-temperature detector calibration, particularly HOLMES. The genuinely new thing is the source concept: an external 280 nm LED driving two 400 nm commercial aluminum foils inside the cryostat, plus a biased grid, giving a switchable electron source whose yield is in principle decoupled from acceleration energy. That is a real step beyond expensive vacuum-sealed photocathodes and CNT field emitters. The qualitative case for electrons is well made: zero signal at 0 V and +300 V, rate scaling quadratically with LED current as expected for photoemission, and pulse energy rising with bias. Seeing 100–300 eV electrons with the HOLMES TES array is also a useful data point for the low-energy electron response literature.\n\nThe soft spots are concentrated in the quantitative claims. The abstract's ≳1 Hz rate and ≳10^-14 e-/gamma efficiency are load-bearing and fragile. They come from a 1D Gaussian fit to the upper row only; the lower row was excluded because one channel had higher noise and a different trigger threshold. That fit gives an x-width but says nothing about the y-profile or y-centering, and the step to a 2D Gaussian is an assumption with no propagated uncertainty. Per-channel rates are tens of mHz, so the integrated rate could plausibly be well below 1 Hz or well above depending on beam shape. The fix is straightforward: more active channels, a collimator or direct rate measurement, and error bars. There is also a mild self-benchmarking loop in the spectral analysis: the model is chosen because it fits Nebula, then the same simulated curves are overlaid on the fit. That is not fatal, because the electron identification does not depend on the endpoint fit. The unresolved escape mechanism from superconducting aluminum is honestly flagged in Section 6 and remains a physics puzzle, but it does not undercut the empirical observation.\n\nOverall, the qualitative result is credible and the source concept is worth developing. The quantitative claims need revision or re-measurement before being quoted. I would send it to peer review rather than desk-reject, with a clear request to address the beam-profile extrapolation and uncertainty propagation. A serious referee should engage.","headline":"A worthwhile prototype demonstration whose qualitative electron-production case is solid, but the headline rate/efficiency numbers rest on a one-row Gaussian extrapolation and should not be taken at face value.","tokens_in":15850,"tokens_out":1903,"would_cite":true,"duration_ms":18594,"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 LED-illuminated aluminum foil produces switchable cryogenic electrons at rates above one per second.","keywords":["electron source","cryogenic detector calibration","photoelectric emission","transition-edge sensors","low-energy electrons","microcalorimeter","LED photocathode","detector calibration"],"falsifier":"Measure the full two-dimensional beam profile by reading all rows with uniform trigger thresholds or by inserting a collimator, and compare the integrated rate with the Gaussian-extrapolated value; an order-of-magnitude discrepancy would invalidate the claimed >=1 Hz rate and >=10^-14 efficiency.","tokens_in":14723,"feed_emoji":"⚡","tokens_out":5667,"duration_ms":51289,"temperature":0.7,"pith_summary":"The paper reports the first working prototype of a compact electron source that can be switched on and off at cryogenic temperatures. It generates electrons by shining a 280 nm LED onto two thin aluminum layers inside a dilution refrigerator, then accelerates them up to 300 V toward an array of microcalorimeters. The observed signals are consistent with electrons arriving at a rate of at least one per second, and the production efficiency is around one electron for every 10^14 photons. This matters because low-temperature detectors currently rely on radioactive sources that cannot be turned off or on field emitters whose yield is tied to energy. A switchable source with independently controllable yield and energy would allow calibration lines to be interleaved with physics data, reducing systematic errors in precision measurements such as neutrino-mass searches.","feed_headline":"LED-lit aluminum emits switchable cryogenic electrons","feed_subtitle":"A photoelectric source that can be pulsed could calibrate low-temperature detectors while data is being taken.","key_machinery":"The mechanism is photoelectric emission from a double layer of 400 nm aluminum foils, illuminated by a 280 nm LED through an optical fiber. The electrons are accelerated by a negative voltage applied to the foils and travel 2.5 cm to a grounded array of microwave-multiplexed transition-edge sensors. The double foil serves as both a photocathode and a light-tight barrier against pinholes. A simple polarity test—signals seen only with negative bias—is the key identity that identifies the particles as electrons, and the count-rate scaling with LED current ties the process to photoemission.","core_discovery":"The central claim is that photoelectric emission from commercial aluminum foils, driven by a fiber-coupled UV LED, is a viable electron production mechanism inside a cryostat at 100 mK. With the foils biased negatively at 100–300 V relative to the grounded detector, single electrons are recorded by transition-edge sensor microcalorimeters; signals appear only under negative bias, confirming they are electrons. The measured rate scales with LED current and the pulse amplitude scales with accelerating voltage, and the inferred source rate and efficiency are >=1 Hz and >=10^-14 electrons per photon. The same setup also demonstrates that TES microcalorimeters can detect directly incident electro","pith_inferences":["The quoted efficiency is per-photon at the photocathode; if the LED-to-foil photon flux is overestimated or the beam profile is broader than the single-row Gaussian fit implies, the true yield could easily be an order of magnitude lower, which would set a demanding requirement on the future electron-multiplier gain.","If yield really is energy-independent, the same device could serve as a tunable monoenergetic electron source for mapping the energy response and nonlinearity of microcalorimeters across the full range of interest.","A direct test would be to replace the current two-foil photocathode with foils of different thickness or different number of layers and observe whether the rate changes; the paper notes this dependence is still unknown.","The surprising escape of electrons from superconducting aluminum suggests the photoemission at mK temperatures may proceed through a mechanism not described by standard mean-free-path arguments; understanding that mechanism could open a route to much higher efficiency."],"forward_implications":["The source can be built from inexpensive, commercially available aluminum and a standard LED, and it operates at the base temperature of a dilution refrigerator without an ultra-high vacuum.","Electron yield is set independently of electron energy: LED current controls the rate, and acceleration voltage controls the energy.","TES microcalorimeters can act as single-electron detectors in the 100–300 eV range, providing a low-energy testbed for these sensors.","By adding an electron multiplier and a conversion target, the source could be developed into a switchable X-ray calibration line above the endpoint of a neutrino-mass spectrum.","Because the LED can be pulsed, calibration data can be collected during short intervals interleaved with physics data, without moving parts."],"fun_headline_variants":["Cryogenic electron gun fired by LED for detector calibration","Switchable cryo electron source uses LED and aluminum foil","LED-driven electron emission at 100 mK calibrates TES detectors","Pulsed electrons from LED-lit aluminum in cryostat for calibration"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The headline rate and efficiency rest on assuming a 2D Gaussian beam shape fitted only to the upper row of the array; if the real beam is wider or asymmetric, those numbers could be off substantially.","fun_headline_variants_meta":{"raw":{"variants":["Cryogenic electron gun fired by LED for detector calibration","Switchable cryo electron source uses LED and aluminum foil","LED-driven electron emission at 100 mK calibrates TES detectors","Pulsed electrons from LED-lit aluminum in cryostat for calibration"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000223,"raw_usage":{"total_tokens":1264,"prompt_tokens":684,"completion_tokens":580,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":428,"completion_tokens_details":{"reasoning_tokens":510}},"tokens_in":428,"tokens_out":580,"duration_ms":5844,"temperature":1.0,"reasoning_tokens":510,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T10:08:19.065049+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the full two-dimensional beam profile by reading all rows with uniform trigger thresholds or by inserting a collimator, and compare the integrated rate with the Gaussian-extrapolated value; an order-of-magnitude discrepancy would invalidate the claimed >=1 Hz rate and >=10^-14 efficiency.","supporting_citations":[],"review_version":1}