{"id":"b5dd3653-307c-4804-a660-537c728ad17f","arxiv_id":"2507.11585","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"First measurements of UV and visible photo- and radio-luminescence of thin ThF4 films show background rates low enough that a small, enriched ThF4 film remains a viable 229mTh nuclear clock target.","lead":"This paper measures how much light thin films of thorium tetrafluoride emit when hit by UV light or alpha particles, then compares those background rates to the expected signal of the 229mTh nuclear isomer. The authors conclude that a small, thin, highly enriched ThF4 film could still serve as a target for a future nuclear clock.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sec. 8 radio-luminescence rescaling uses an incorrect mean alpha path ('half of the film') for internal decays; the true mean path in a 50 nm slab is far longer, so the claimed 0.2 Hz background is likely underestimated.","rationale":"The reader's weakest assumption correctly identified the unverified scaling from thick amorphous commercial films to a thin crystalline enriched film, but did not isolate the specific mean-path error inside the radio-luminescence rescaling. This error is load-bearing because it directly multiplies the most important continuous background in the viability projection, and the factor is not a small correction: for a thin slab the mean exit distance for an isotropic internal source is much larger than half the thickness, so the 0.2 Hz estimate is likely too low by a factor of several. The paper's measurements and Monte Carlo acceptance work are careful and transparent, and the conclusion that ThF4 is a promising matrix may survive, but the quantitative statement that backgrounds 'do not exceed acceptable limits' is not established without a correct path-length treatment or a direct measurement on a thin film. The Eq. 7 versus 1.1 Hz signal-rate inconsistency is real and should be corrected, but it is less central because a higher signal would make the background situation more favorable; it does not rescue the background-path error. The appropriate verdict remains CONDITIONAL: the paper needs revision and, ideally, a dedicated thin-film measurement before the viability claim is accepted as stated. Since the reader already assigned CONDITIONAL, my read does not change the verdict, though it sharpens the reason.","tokens_in":15114,"tokens_out":17262,"duration_ms":218490,"concrete_test":"Run a Geant4 or analytic calculation of the mean alpha path length inside a 50 nm, 1 cm2 ThF4 slab for uniform, isotropic, internal 5 MeV alpha emission, and rescale the measured UV radio-luminescence probability (about 0.0019 per external alpha crossing 200 nm) using that mean path. If the mean path is ~0.25–1 μm instead of 25 nm, the internal background becomes ~1–5 Hz and the Sec. 8 'acceptable limits' claim fails. Ideally, also measure radio-luminescence from a 50 nm ThF4 film with implanted 229Th, or from a 50 nm film irradiated by a collimated alpha beam in the same geometry, to test the scaling directly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central viability claim in Sec. 8 rests on rescaling the measured UV radio-luminescence of a 200 nm film (0.7 Hz under the external 241Am source, Sec. 6.1) to an internal 1 kBq, 50 nm, 1 cm2 229Th-doped film, obtaining 0.2 Hz. The text justifies this by stating that external alphas cross the entire film while 'intrinsic activity αs, on average, will cross only half of it.' That geometric assumption is incorrect for isotropic internal sources in a thin slab. For a point inside a slab, the distance to exit in a random direction is z/|cosθ| (or (t−z)/|cosθ|); averaging over uniform depth z and isotropic direction yields a mean path that diverges logarithmically for an infinite slab and is cut off only by lateral dimensions or alpha range. For a 50 nm film with 1 cm2 area and a ~10 μm alpha range, the mean path is of order t·ln(R/t) ≈ 50 nm × ln(10 μm / 50 nm) ≈ 0.27 μm, not 25 nm. Thus the internal radio-luminescence rate could be roughly 1–5 Hz rather than 0.2 Hz. Since Sec. 7 sets the acceptable continuous background near 0.1 Hz for the nominal 6-signal projection, even a 1 Hz background exceeds the quoted limit and would require far more repetition than implied. Separately, Eq. 7 evaluates to about 6.5 Hz when n = 1.826 is inserted, while the text uses 1.1 Hz; if the signal is higher, the limit loosens, but the paper's own stated comparisons use the lower value, and neither the signal normalization nor the background path-length scaling has been verified on a 50 nm crystalline film.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports measurements of VUV-excited photo-luminescence and 241Am alpha-induced radio-luminescence from commercially produced amorphous ThF4 films (200 nm on ZnSe, 300 nm on Si), using a slow charge-sensitive PMT readout with separate VUV and visible detectors. The authors find that UV luminescence rates are comparable to PMT dark counts (0.23 Hz photo-luminescence from the 200 nm film; 0.70 ± 0.06 Hz UV radio-luminescence from the 200 nm film), derive rough light yields (≈0.09 photons/keV at 180 nm in the UV), and use Geant4 acceptance simulations to project signal and background rates for a hypothetical 50 nm, 1 cm2, 1 kBq 229Th-doped ThF4 target. They conclude that the measured backgrounds are significant but acceptable for detecting roughly six 229mTh decay events in 360 s.","tokens_in":15515,"tokens_out":9137,"duration_ms":101000,"significance":"If the central projection were valid, this would be a useful first characterization of ThF4 thin-film luminescence for nuclear-clock applications, and the measured film-thickness scaling (1.46 ± 0.16 for a 1.5 thickness ratio) is a nice internal consistency check. The paper reports reproducible raw rates and a conservative readout scheme, and it gives Geant4-based acceptance estimates. However, two load-bearing quantitative steps—the numerical evaluation of Eq. (7) and the geometric rescaling of internal radio-luminescence in Sec. 8—are not correct as written; both need revision before the viability conclusion can be relied on.","major_comments":[{"comment":"Equation (7) is not evaluated as stated. With the quoted flux dNγ/dtdλ(λ0) = 1.6×10^13 ph/(cm2 s nm) and n(150 nm) ≈ 1.826, n^3 ≈ 6.09, the expression 6.7×10^-14 nm × n^3 × dNγ/dtdλ gives approximately 6.5 Hz, not 1.1 Hz. The subsequent numbers 217 signal events, 6 detected events, and the 0.017 Hz average signal rate in Sec. 7 are all derived from the 1.1 Hz value; until this arithmetic is corrected (or the flux/units are clarified), the signal-to-background ratios quoted in Sec. 8 are internally inconsistent.","section":"Sec. 7, Eq. (7)"},{"comment":"The rescaling of the measured external-alpha radio-luminescence to internal 229Th activity uses an incorrect mean path. For an isotropic internal source in a 50 nm slab, the average path length in the film before exiting is not t/2 = 25 nm: the distance to the surface in direction θ is z/|cosθ| or (t − z)/|cosθ|, and for a slab with lateral size large compared with the alpha range R the mean path is of order t ln(R/t). With t = 50 nm and R ≈ 10 µm this is ≈0.27 µm, more than ten times the value assumed in the text. The resulting internal UV radio-luminescence background is therefore plausibly several hertz rather than 0.2 Hz, exceeding the <0.1 Hz continuous-background budget set in Sec. 7. The conclusion that the internal radio-luminescence is 'within an acceptable range' needs to be re-derived with the correct path-length distribution or with a Monte Carlo that tracks internal decays.","section":"Sec. 8, radio-luminescence bullet"},{"comment":"The final viability statement is supported only by a multi-step extrapolation from the measured samples to an unmeasured target: the proposed 50 nm crystalline 229Th-doped film was not fabricated or tested; the frustrated-TIR acceptance recovery at 50 nm (about 7%) is estimated, not measured; and the downward scaling of the 200–300 nm amorphous-film backgrounds assumes strict proportionality to film volume and area, with negligible luminescence from the intermediate coating layers beneath ThF4. Since the margins after scaling are small (e.g., the rescaled 0.2 Hz internal radio-luminescence versus the 0.1 Hz budget), these assumptions are load-bearing. At minimum, the conclusions should be reworded as an indicative estimate, and a measurement on a 50 nm film—ideally with 229Th doping—should be listed as required validation.","section":"Sec. 8 and Sec. 7"}],"minor_comments":[{"comment":"The caption of Fig. 11 describes the fitted curve as having a maximum amplitude of 0.23 Hz, but Sec. 4.1 states that the ThF4+Si sample peaks at about 0.35 Hz; the caption appears to be copied from Fig. 10 and should be corrected.","section":"Figs. 10 and 11"},{"comment":"The notation '0.23 Hz/0.69' and '1.4 kHz ×0.70/0.70' is unclear; presumably these are rates corrected by a 69–70% signal-selection efficiency, but the efficiency correction should be written explicitly rather than as a slash or ratio.","section":"Secs. 4.1 and 4.2"},{"comment":"The statement that the fraction of photo-diode current above 200 nm was 'estimated to be 55%, which is significantly higher than the observed 33%' is confusing: if 55% of the current is above 200 nm, one would expect more than 33% in air; please clarify what accounts for the shortfall.","section":"Sec. 2.1"},{"comment":"The sentence 'we used scintillation parameters adapted to the measurement results' is vague; since the light yield is derived from the measured rate and the simulated acceptance, please clarify that this is not a fit that could create a circular determination.","section":"Sec. 5"},{"comment":"Minor wording issue: 'the using of a small and thin ThF4 film' should read 'the use of a small and thin ThF4 film'.","section":"Sec. 8"}],"recommendation":"major_revision","confidential_remarks":"Confidential to the editor: The paper fits the journal's scope. The main risk is that the central viability claim in Sec. 8 is currently supported by an arithmetic error in Eq. (7) and an incorrect geometric assumption about internal alpha path lengths; a revised version that fixes these and softens the extrapolation would be acceptable. There is no novelty or citation concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The useful part first: this is the first quantitative characterization of photo- and radio-luminescence of thin ThF4 films. The experimental work is careful: the background subtraction between sample sides is sensible, the thickness ratio check (1.46 ± 0.16 vs 1.5) is a good control, and the observation of a ~400 s UV luminescence lifetime is new and relevant. The apparatus description is detailed enough to reproduce. For anyone designing a 229mTh solid-state target, the raw rates are a useful input.\n\nThe soft spots are in the projection, not the measurement. The Sec. 8 rescaling from the measured external-alpha radio-luminescence to an internal 1 kBq, 50 nm, 1 cm2 film uses the claim that internal alphas cross \"on average only half\" the film. That is wrong for isotropic emission in a thin slab. The mean path before exit is of order t ln(R/t) with R the alpha range or lateral size; for t = 50 nm and R ~ 10 μm this is about 300 nm, an order of magnitude larger than the 25 nm used. The internal radio-luminescence background then becomes several Hz, not 0.2 Hz, and that is above the 0.1 Hz budget the paper itself sets for the 6-signal detection. This is the main issue.\n\nSecond, Eq. (7) does not evaluate to 1.1 Hz when you plug in n = 1.826 and the stated flux of 1.6e13 ph/(cm2 s nm); you get about 6.5 Hz. The text's signal estimate and the resulting 6 events in 360 s are based on the lower number. This is an internal numerical inconsistency; it happens to favor the authors, but it has to be fixed and the projections recomputed.\n\nMinor points: the photo-luminescence fits have no error bars on amplitude or lifetime; the visible ThF4+Si data are excluded post hoc with a plausible contamination argument; and the step from thick amorphous commercial films to a thin crystalline enriched film is unverified. The authors flag the last one, but it remains a leap.\n\nWho it's for: the solid-state 229mTh community. The raw measurement is worth having; the projection section is not. I'd send it to peer review with the expectation of major revision: require the authors to correct the path-length scaling, make the signal calculation self-consistent, and present the viability claim with the corrected numbers. If the corrected radio-luminescence background is indeed ~2 Hz, the conclusion changes from \"acceptable\" to \"needs a different geometry or more enrichment\". The experimental core is solid enough to deserve that work.","headline":"First ThF4 luminescence data look solid, but the background projection in Sec. 8 uses a wrong mean path for internal alphas and the signal estimate has an unaddressed factor-of-6 discrepancy.","tokens_in":16051,"tokens_out":10293,"would_cite":true,"duration_ms":124409,"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":"Thin ThF4 films are quiet enough in the UV to serve as a 229mTh nuclear-clock target.","keywords":["Th-229m isomer","ThF4 thin films","photo-luminescence","radio-luminescence","nuclear clock","vacuum ultraviolet","internal conversion suppression","background characterization"],"falsifier":"Measure UV photo- and radio-luminescence of an actual 50 nm ThF4 film, crystalline and if possible 229Th-doped, under the same lamp and alpha-source geometry. If the UV photo-luminescence does not drop by roughly the predicted factor of 20 relative to the 200 nm film, or if the radio-luminescence at 1 kBq exceeds about 0.1 Hz in 360 s windows, the Section 8 background budget fails.","tokens_in":14923,"feed_emoji":"⚛️","tokens_out":6698,"duration_ms":76676,"temperature":0.7,"pith_summary":"The paper sets out to answer a practical question for a thorium nuclear clock: does a thin film of ThF4 glow too much, under UV light or under alpha irradiation, for the faint decay of the 229mTh isomer to be seen? It reports the first measurements of photo- and radio-luminescence from 200-nm and 300-nm ThF4 films and finds that in the UV range both sit at roughly the level of the photomultiplier dark count. Scaling those rates to a 1 cm2, 50 nm film loaded with 1 kBq of 229Th, the paper concludes that the UV photo-luminescence falls below the expected isomer signal and the internal radio-luminescence background becomes about 0.2 Hz, which is acceptable for detecting roughly six signal events in 360 s. The measurements are offered as evidence that a small, thin, highly enriched ThF4 target is a viable route to observing 229mTh decay and testing the refractive-index dependence of its lifetime.","feed_headline":"Thin ThF4 films keep 229mTh backgrounds acceptable","feed_subtitle":"UV glow sits near PMT dark noise; a 50 nm film should bring radio-luminescence down to 0.2 Hz and make isomer decay visible.","key_machinery":"The load-bearing element is a single-photoelectron counting measurement in two spectral bands. The authors illuminate a ThF4 film with a vacuum-ultraviolet lamp, move the sample to a photomultiplier station, and count pulses in an ADC window from one-third to four-thirds of the single-photoelectron amplitude, isolating faint UV luminescence from a 0.1 Hz dark background. For radio-luminescence they place an alpha source above the sample and subtract substrate-side measurements, using Monte Carlo acceptance simulations to convert observed rates into light yields. The argument that carries the conclusion is a volume-scaling law: luminescence background is proportional to film thickness, while the 229mTh excitation rate is proportional to the areal density of 229Th, so thinning the film and raising enrichment improves signal to background. The expected signal uses the resonant photo-excitation cross section with 1 kBq/cm2 of 229Th and the n3-dependent isomer width.","core_discovery":"On its own terms, the paper's discovery is that ThF4 thin films are quiet enough in the UV to serve as a 229mTh host. Comparing UV luminescence rates measured on the ThF4 side and the substrate side of two commercial optics, the authors isolate 0.70 ± 0.06 Hz (200 nm) and 1.03 ± 0.07 Hz (300 nm) of alpha-induced UV radio-luminescence, whose ratio matches the 1.5 thickness ratio, and a UV photo-luminescence of about 0.23 Hz decaying with a ~400 s lifetime. They then argue that because the background scales with film volume while the 229mTh signal is set by the surface density of 229Th, a 1 cm2, 50 nm film can cut these backgrounds by a factor of about 20 while enrichment keeps the signal at a ~1.1 Hz excitation rate, yielding about six detected decays in 360 s at a 0.027 Hz peak rate. The conclusion is the central sentence of Section 8: the measured backgrounds are significant but do not exceed acceptable limits. The same data also single out ThF4 as a host whose refractive index near 150 nm is far enough from previously tested crystals to test the predicted n3 scaling of the isomer linewidth.","pith_inferences":["If the 400 s UV photo-luminescence decay shares a similar timescale with the expected isomer decay, the volume-scaling assumption should be checked against films of several thicknesses; a surface or defect origin of the luminescence would break that assumption.","The 50 nm film changes the optical boundary condition: the paper estimates that frustrated total internal reflection recovers about half the acceptance, but a sub-wavelength film could exhibit interference or waveguiding effects that alter both acceptance and the apparent decay rate.","A natural extension would be to measure luminescence as a function of film thickness and crystallinity, directly testing the linear volume scaling and separating bulk scintillation from substrate or interface contributions."],"forward_implications":["A 1 cm2, 50 nm ThF4 film with high 229Th enrichment should show UV photo-luminescence below the expected 229mTh signal rate, making the isomer decay visible in a few hundred seconds.","The internal alpha radio-luminescence of the proposed film is estimated at 0.2 Hz, requiring repeated long runs for statistical significance rather than a single measurement.","Visible luminescence of ThF4 is high enough to serve as a veto channel, with negligible probability of accidental coincidence with the UV signal.","The same apparatus can be used directly to search for 229mTh excitation and decay in ThF4, with a natural-Th control sample needed to separate the 400 s luminescence from the isomer lifetime.","A successful ThF4 target would provide a solid-state test of the predicted n3 enhancement of the 229mTh radiative width, since ThF4's refractive index differs strongly from previously tested crystals."],"supporting_citations":[{"why":"Fixes the 8.3 eV transition energy that defines the signal and the 150 nm wavelength used in the rate estimates.","marker":"[1]"},{"why":"Motivates the nuclear clock application that the background budget is meant to enable.","marker":"[2]"},{"why":"Provides a previous solid-state host where internal conversion is suppressed, serving as the larger-crystal baseline for backgrounds.","marker":"[9]"},{"why":"Another host with observed radiative decay, against which ThF4 backgrounds and n3 scaling are compared.","marker":"[10]"},{"why":"Establishes the 10.2 eV band gap that suppresses internal conversion, a premise for using ThF4.","marker":"[13]"},{"why":"Supplies the ThF4 refractive index that sets the predicted isomer linewidth and the total internal reflection acceptance correction.","marker":"[14]"},{"why":"Used to compute the photomultiplier acceptance in the Monte Carlo simulations for both photo- and radio-luminescence.","marker":"[24]"},{"why":"Gives the theoretical matrix element behind the n3-scaled radiative width used in the 229mTh signal estimate.","marker":"[26]"}],"fun_headline_variants":["Thin ThF4 films quiet UV noise for 229mTh","ThF4 films: UV glow near dark noise, 20x less","Thin ThF4 films make 229mTh search viable","ThF4 thin films: 20x background cut for nuclear clock"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The projection assumes the luminescence measured in thick commercial amorphous films scales linearly with film volume down to a 50 nm crystalline film, and that the 229mTh signal can be held fixed by enrichment; neither a thin film, a crystalline film, nor a 229Th-doped film was measured.","fun_headline_variants_meta":{"raw":{"variants":["Thin ThF4 films quiet UV noise for 229mTh","ThF4 films: UV glow near dark noise, 20x less","Thin ThF4 films make 229mTh search viable","ThF4 thin films: 20x background cut for nuclear clock"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000209,"raw_usage":{"total_tokens":1417,"prompt_tokens":964,"completion_tokens":453,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":580,"completion_tokens_details":{"reasoning_tokens":374}},"tokens_in":580,"tokens_out":453,"duration_ms":6192,"temperature":1.0,"reasoning_tokens":374,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:13:15.928512+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure UV photo- and radio-luminescence of an actual 50 nm ThF4 film, crystalline and if possible 229Th-doped, under the same lamp and alpha-source geometry. If the UV photo-luminescence does not drop by roughly the predicted factor of 20 relative to the 200 nm film, or if the radio-luminescence at 1 kBq exceeds about 0.1 Hz in 360 s windows, the Section 8 background budget fails.","supporting_citations":[{"cited_title":"Seiferle, et al., Energy of the 229Th nuclear clock transition, N ature 573 (2019) 243","cited_arxiv_id":null,"evidence_quote":"Fixes the 8.3 eV transition energy that defines the signal and the 150 nm wavelength used in the rate estimates."},{"cited_title":"Tkalya, et al., Processes of the nuclear isomer 229mTh(3/2+, 3.5 +- 1.0 eV) resonant excitation by optical photons, Phys","cited_arxiv_id":null,"evidence_quote":"Motivates the nuclear clock application that the background budget is meant to enable."},{"cited_title":"Tiedau, et al., Laser excitation of the Th-229 nucleus, Phys","cited_arxiv_id":null,"evidence_quote":"Provides a previous solid-state host where internal conversion is suppressed, serving as the larger-crystal baseline for backgrounds."},{"cited_title":"Kraemer, et al., Observation of the radiative decay of the 22 9Th nuclear clock isomer, Nature 617 (2023) 706","cited_arxiv_id":null,"evidence_quote":"Another host with observed radiative decay, against which ThF4 backgrounds and n3 scaling are compared."},{"cited_title":"Gouder, et al., Measurements of the band gap of ThF 4 by electron spectroscopy techniques, Phys","cited_arxiv_id":null,"evidence_quote":"Establishes the 10.2 eV band gap that suppresses internal conversion, a premise for using ThF4."},{"cited_title":"Heitmann, E","cited_arxiv_id":null,"evidence_quote":"Supplies the ThF4 refractive index that sets the predicted isomer linewidth and the total internal reflection acceptance correction."},{"cited_title":"Agostinelli, et al., Geant4—a simulation toolkit, Nucl","cited_arxiv_id":null,"evidence_quote":"Used to compute the photomultiplier acceptance in the Monte Carlo simulations for both photo- and radio-luminescence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the theoretical matrix element behind the n3-scaled radiative width used in the 229mTh signal estimate."}],"review_version":1}