{"id":"4cad22f5-e8e0-4baf-a63b-40f3692c875d","arxiv_id":"2511.12328","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Absolute photon-detection efficiency of the X-ARAPUCA for DUNE's vertical-drift module is 3.1–4.7%, and removing dichroic filters improves it by up to 18% (11% in the same-substrate comparison).","lead":"DUNE's photon-detector team measured how efficiently a new light-collecting device (X-ARAPUCA) detects vacuum-ultraviolet light in liquid argon, reporting absolute efficiencies of 3.1–4.7% depending on voltage and design, and a filter-free variant that performs better. The results validate the detector against DUNE's physics requirement of >3% efficiency and support a design change to remove a dichroic filter from the baseline.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absolute PDE scale rests on one external VUV4 SiPM calibration plus an LXe angle-transfer; a ~19% systematic there would drop the filtered baseline below the 3% DUNE threshold, and no in-paper test pins it down.","rationale":"The paper's relative comparisons and repeatability are strong. The two-method agreement and the same-substrate DS comparison independently support the 11% filter-removal effect; the 18% SS number is partly confounded by a substrate change (P.E. vs ZAOT), but that weakens wording, not the design conclusion. The most load-bearing unresolved element is the absolute scale. Eq. 2 uses ε_ref = 12.1±1.1% as a multiplicative input; the paper's error budget says it dominates. The reference calibration is external (ref [20]) and the angle correction is imported from LXe (ref [22]) at a different wavelength; the paper does not test either. The simulation method is the only independent absolute check, but its bending correction (f_bending=0.58±0.08) was derived after data-taking, is large, and is acknowledged in the text as assumption-dependent. Thus the cross-check cannot rule out a >10% scale error. A 19% scale reduction would drop the filtered baseline below the DUNE 3% threshold; noDF configurations keep a margin until ~33%, so the filter-free recommendation is less vulnerable than the absolute efficiencies. This is exactly the reader's weakest assumption, so agreement is 'agree'. The appropriate verdict remains CONDITIONAL: the internal logic is sound and relative claims are well supported, but the absolute compliance claim should not be treated as definitive until the reference scale is independently demonstrated. If the proposed calibration measurement agrees with 12.1% within ~5%, I would accept the paper.","tokens_in":18610,"tokens_out":10185,"duration_ms":92358,"concrete_test":"Measure the PDE of the same HPK VUV4 reference sensors in LAr at 127 nm and 87 K as a function of incidence angle (0°, 30°, 45°, 60°, 75°) against a calibrated photodiode or NIST-traceable source, and compute the angle-weighted ε_ref from the simulated photon angular distribution. Recompute Eq. 2 with this ε_ref and propagate through Table 9. If the weighted ε_ref differs from 12.1% by >10% (or if DF-XA at 4.5 V falls below 3%), the absolute scale needs renormalization; if it agrees within ~5%, the central claim stands.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Every reference-method value in Table 9 (headline 3.7±0.3% and 4.5±0.4%) is proportional to ε_ref in Eq. 2. ε_ref is the previously measured VUV4 PDE, (12.69±1.12)% at 127 nm, 87 K, 4 V OV [20], then rescaled to 12.1±1.1% using a single average incidence angle (75°) from [22]—a liquid-xenon measurement at a different VUV wavelength. The error budget lists this calibration as the dominant 8.7% term, and the paper performs no independent check of the absolute scale. The simulation method could have supplied one, but its f_bending=0.58±0.08 correction (Sec. 3.2) is post-hoc, characterized after data-taking, and carries ~14% uncertainty; the 3.1±0.5 vs 3.7±0.3 agreement therefore constrains the reference scale only loosely. A ~19% downward shift in ε_ref would put the filtered baseline at 4.5 V below the 3% DUNE requirement; the noDF configurations would still pass until ~33%. The relative improvements in Table 10 are largely calibration-independent, so the design recommendation is less exposed—what is exposed is the absolute compliance claim and every absolute PDE quoted for DUNE PDS.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports laboratory measurements of the absolute photon-detection efficiency (PDE) of the X-ARAPUCA photosensor proposed for DUNE's Vertical Drift far detector. Two independent methods are used: a 'reference method' at CIEMAT, which normalizes the XA signal to two calibrated HPK VUV4 SiPMs and a Geant4 geometrical factor, and a 'simulation method' at INFN Naples, which derives the absolute number of incident photons from a Geant4 simulation with liquid-argon purity and light-guide bending corrections. For the baseline single-sided XA at 4.5 V overvoltage, the two methods give 3.7±0.3% and 3.1±0.5%, respectively, compatible within ~1σ. Several alternative configurations are also measured; the configuration without dichroic filters reaches 4.5±0.4% at 4.5 V, and the paper claims improvements of up to 18% (single-sided) or 11% (double-sided) relative to the filtered baseline. Based on these results, the authors recommend removing the dichroic filter from the DUNE VD baseline design. An alternative WLS light-guide geometry shows no significant absolute-PDE improvement but a flatter position response.","tokens_in":19005,"tokens_out":9714,"duration_ms":84068,"significance":"If the absolute values are correct, this is an important result for the DUNE photon-detection system: it indicates that the VD baseline XA meets the >3% PDE requirement and that the design can be simplified by removing the dichroic filters. The paper's strengths include two independent cryogenic setups, an explicit error budget, repeated measurements with demonstrated stability, and the confirmation of a simulation-driven prediction about the dichroic-filter loss. The relative noDF-vs-DF comparisons are largely insensitive to the external SiPM calibration, which makes the design recommendation more robust than the absolute numbers. However, the absolute PDE scale rests on a single external calibration and a liquid-xenon angle-transfer, the headline 18% improvement is not a same-substrate comparison, and the simulation-method cross-check depends on a post-hoc bending correction. These caveats limit the strength of the absolute compliance claim and of the claimed cross-validation.","major_comments":[{"comment":"The absolute PDE scale of the reference method is proportional to ε_ref, the previously measured HPK VUV4 PDE of (12.69±1.12)% at 127 nm, 87 K, 4 V OV [20], rescaled to 12.1±1.1% using an average incidence angle from a liquid-xenon measurement [22]. The error budget in §5.1 lists this external calibration as the dominant 8.7% term, and the paper performs no independent check of this scale. The simulation method could in principle cross-check it, but its dominant correction f_bending=0.58±0.08 is post-hoc (see below), so the 3.7±0.3 vs 3.1±0.5 agreement cannot exclude a systematic shift of ~19% in ε_ref, which would put the filtered baseline below the DUNE >3% threshold. Please quantify this sensitivity and either provide an independent anchor or temper the absolute compliance claim.","section":"§2.1, §2.2, Eq. (2); §5.1 error budget"},{"comment":"The headline 18% noDF-vs-DF improvement for the single-sided configuration compares Configuration 1 (DF, ZAOT substrate) with Configuration 3 (noDF, P.E. substrate). These differ by two variables, so the improvement cannot be attributed solely to filter removal. The controlled same-substrate comparison is the double-sided Configuration 2 vs 4 (both ZAOT), which gives 11%. The abstract and conclusion should quote the DS comparison as the primary evidence for removing the DF, or report a substrate-corrected SS number; otherwise 'up to 18%' overstates the filter effect.","section":"§2.3 Table 4; §5.3 Table 10"},{"comment":"The simulation method relies on a bending correction f_bending=0.58±0.08 characterized after data-taking, with the text acknowledging the bending 'might have been more dramatic during the data taking.' This is a post-hoc normalization of the INFN data; together with its ~14% uncertainty it weakens the claimed cross-check of the reference method. The authors should present the simulation-method results with and without this correction, or treat the cross-check as qualitative rather than as an independent validation of the absolute scale.","section":"§3.2, Eq. (8)"},{"comment":"Eq. (3) states f_geo = MC_ref/MC_XA = 0.0465±0.0007, but the values given in Table 3 and the text are MC_ref=603±7 and MC_XA=25,900±238, whose ratio is 0.0233, not 0.0465. Since the reference-method PDE is proportional to f_geo, this factor-of-two discrepancy directly affects all absolute values in Table 9. Please clarify the definition (e.g., per-SiPM vs summed, or an additional area factor) or correct the value; if the raw MC ratio is the intended f_geo, the absolute PDEs would shift by a factor of two.","section":"§2.2, Eq. (3), Table 3"}],"minor_comments":[{"comment":"The abstract reads '4.5±4%' but the correct value is 4.5±0.4%; the decimal point is missing.","section":"Abstract"},{"comment":"The substrate designation 'P.E.' is never defined. Please spell out the material and, if relevant, note whether it is the same substrate as ZAOT apart from the dichroic coating.","section":"Table 4"},{"comment":"Typo: 'uncertaty' should be 'uncertainty'.","section":"§1.3, Eq. (1)"},{"comment":"When combining the three overvoltage measurements in Table 10, the statement that 'the overvoltage choice should have no impact on the relative PDE' is an assumption; please justify it or show that the three rows are compatible within their uncertainties before averaging.","section":"§5.3"},{"comment":"The figure of merit 'PDE times Surface over #SiPM' (26 vs 83) is not defined in this paper; please define it or give a reference for this metric.","section":"§6"},{"comment":"Reference [17] is cited as 'Manuscript in preparation'; please replace it with a citable publication or, if unavailable, note that the result is preliminary.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of EPJC and the collaboration authorship is appropriate. The most urgent technical point is the apparent inconsistency between Eq. (3) and the numbers in Table 3; if the factor-of-two discrepancy is a typo, it is easily fixed, but if it reflects an incorrect f_geo, the absolute PDE values change materially. The external-calibration reliance is a normal limitation for this type of measurement, but the authors should be explicit about how much the absolute compliance claim depends on it. The substrate confound in the 18% improvement should also be corrected in the abstract and conclusions. No concerns about the citation pattern or novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is the first LAr characterization of the VD X-ARAPUCA, and it's a genuinely useful result. The two independent methods (reference SiPM and simulation) give 3.7±0.3% versus 3.1±0.5% at 4.5 V OV, consistent within uncertainties. The biggest design message — removing the dichroic filters helps — is credible, since the same-substrate DS comparison shows an 11% gain and the no-DF configurations were motivated by simulation before the measurement. That's a real prediction, not a retrofit.\n\nThe error budget is honestly reported. The reference method's dominant uncertainty is the previously measured VUV4 SiPM PDE, and the paper doesn't hide that. The stability/repeatability checks over a year are also a plus.\n\nThe soft spots are real but manageable. First, the absolute PDE scale is not independently pinned in this paper. Every reference-method value is proportional to ε_ref from [20], with the angle correction taken from liquid xenon. A ~19% downward shift in that calibration would put the filtered baseline below DUNE's 3% requirement. The simulation method could have provided an independent anchor, but its f_bending = 0.58±0.08 correction was characterized after data-taking and carries ~14% uncertainty, so the agreement between methods is a loose constraint. This is not a fatal flaw — the authors flag it and the relative comparisons are largely calibration-independent — but a sensitivity statement or a cross-check with a second reference sensor would strengthen the compliance claim.\n\nSecond, the headline 'up to 18%' improvement for the single-sided configuration confounds filter removal with a substrate change (P.E. vs ZAOT). The DS comparison, same ZAOT substrate, gives 11%, which is the cleaner number. The paper does show both, but the abstract should be reworded.\n\nThird, the abstract's '4.5±4%' is a typo; the body says 4.5±0.4%.\n\nNo data or code is released. That's not unusual for a detector paper, but the MC parameter values would help others reproduce the geometry correction.\n\nWho is this for? Anyone working on DUNE's PDS, X-ARAPUCA development, or absolute PDE metrology in LAr. It deserves a serious referee. I'd accept it with minor revisions: fix the typo, re-frame the headline improvement, and add one paragraph discussing how robust the absolute scale is to the reference SiPM calibration. The design recommendation to remove the DFs is safe either way.","headline":"Solid two-method absolute PDE measurement for DUNE's VD X-ARAPUCA; the no-filter gain is real, but the absolute scale leans on one external SiPM calibration.","tokens_in":19767,"tokens_out":3047,"would_cite":true,"duration_ms":28304,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["29.40.Mc","85.60.Gz","95.55.Vj"],"model":"deepseek-v4-flash","headline":"The X-ARAPUCA photon detector for DUNE's Vertical Drift module reaches an absolute efficiency of 3.7% at 4.5 V overvoltage, enough for the experiment's physics goals.","keywords":["X-ARAPUCA","photon detection efficiency","liquid argon","DUNE vertical drift","silicon photomultiplier","dichroic filter","wavelength shifter","scintillation light"],"falsifier":"Re-measure the reference SiPM's PDE directly at 127 nm in liquid argon at the actual incidence angles of this setup (~75°) rather than using the liquid-xenon-based angle correction; if the resulting reference PDE differs from the assumed 12.1%, the X-ARAPUCA's absolute PDE values in Table 9 shift by that factor. Alternatively, run the reference-method comparison with a different calibrated reference device (e.g., a second SiPM model or a photodiode-based detector) and check that the XA PDE remains 3.7% at 4.5 V overvoltage.","tokens_in":18480,"feed_emoji":"⚛️","tokens_out":5273,"duration_ms":44360,"temperature":0.7,"pith_summary":"This paper establishes the absolute photon-detection efficiency (PDE) of the X-ARAPUCA light collector proposed for DUNE's Vertical Drift far detector, measured directly in liquid argon under cryogenic conditions. The baseline single-sided design reaches a PDE of 3.7 ± 0.3% at 4.5 V overvoltage, above the >3% value DUNE needs to meet its 20 PE/MeV light-yield requirement; a second, independent simulation-based method gives a compatible 3.1 ± 0.5%. The paper also shows that removing the dichroic filters improves efficiency by up to 18% because the filters' transmittance losses outweigh their photon-recycling benefit, and recommends dropping them from the baseline. These results matter because they validate a simpler and cheaper photon-detection design while demonstrating a methodology that cancels common biases by comparing the device under test against a calibrated reference sensor.","feed_headline":"X-ARAPUCA tops DUNE efficiency bar at 3.7%","feed_subtitle":"Two independent liquid-argon tests agree, and cutting the dichroic filter adds up to 18% more light.","key_machinery":"The measurement rests on the X-ARAPUCA device itself—a wavelength-shifting light guide (PMMA doped with a chromophore) with an external p-terphenyl coating that converts 127 nm argon scintillation light to ~350 nm, traps re-emitted visible light by total internal reflection, and delivers it to 160 SiPMs. The efficiency is extracted through two independent methods. The reference method compares the XA's detected photoelectrons against those of a calibrated reference SiPM (known PDE of 12.69 ± 1.12% at 127 nm, 87 K, 4 V overvoltage) using a Geant4-simulated geometrical factor that cancels global light-yield biases; the simulation method instead computes the absolute number of incident photons","core_discovery":"The central claim is that the X-ARAPUCA baseline configuration for DUNE's Vertical Drift module meets the experiment's photon-detection requirements: the reference method measures an absolute PDE of (3.7 ± 0.3)% at 4.5 V overvoltage, and the independent simulation method measures (3.1 ± 0.5)%, with the two agreeing within uncertainties. The paper further claims that the dichroic filters included in the baseline design to re-reflect escaping photons actually reduce efficiency: removing them produces an 18% improvement for the single-sided configuration and an 11% improvement for the double-sided one, with the best configuration reaching (4.5 ± 0.4)% at 4.5 V. On this basis the authors recomme","pith_inferences":["Because all absolute PDE values inherit the reference SiPM's calibration, a systematic error in that calibration would shift every number proportionally; the relative comparisons (e.g., the 18% gain from removing dichroic filters) would survive such an error.","The 18% improvement attributed to dichroic-filter transmittance losses suggests a concrete design lever: further gains may come from increasing the reflectivity of the non-active surfaces (the VIKUITI film) or from dichroic filters with sharper cut-offs and higher transmittance, rather than from removing the filters alone.","The flatter position response of the alternative light guide implies that a configuration with slightly lower mean PDE could still yield better physics performance if uniformity matters more than peak efficiency; a detector-level light-yield simulation could test this.","A natural testable extension is to reproduce the absolute measurement with a third, independent calibration—for instance a calibrated photodiode or a different reference SiPM batch—to verify the 3.7% central value without relying on the liquid-xenon angle-dependence correction."],"forward_implications":["DUNE's Vertical Drift photon detection system can meet the 20 PE/MeV light-yield requirement with the current X-ARAPUCA design: the measured 3.7% PDE exceeds the needed >3%.","The dichroic-filter-free X-ARAPUCA becomes the recommended baseline; it is simpler and cheaper while improving efficiency by up to 18%.","Single- and double-sided X-ARAPUCAs are compatible in efficiency, so cathode-mounted double-sided devices do not pay an efficiency penalty.","The two-method cross-check establishes a repeatable procedure for qualifying future photon-detector designs for DUNE against an absolute PDE target.","No efficiency gain comes from the alternative 24 mg/kg, 5.5 mm light guide, though its flatter position response could improve event reconstruction uniformity."],"fun_headline_variants":["DUNE X-ARAPUCA hits 3.7% PDE, filter-free design gains 18%","Removing dichroic filters boosts DUNE X-ARAPUCA efficiency by 18%","X-ARAPUCA for DUNE: 3.7% PDE, filters cost 18% light","Liquid argon test: X-ARAPUCA without filters beats baseline by 18%","DUNE photon detector: X-ARAPUCA efficiency 3.7%, filters worse"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The absolute efficiency numbers depend on the previously measured efficiency of the reference SiPMs (12.69 ± 1.12% at 127 nm, 87 K, 4 V overvoltage) and on transferring its angle dependence from liquid xenon to liquid argon; if that external calibration is systematically wrong, every absolute PDE value shifts proportionally.","fun_headline_variants_meta":{"raw":{"variants":["DUNE X-ARAPUCA hits 3.7% PDE, filter-free design gains 18%","Removing dichroic filters boosts DUNE X-ARAPUCA efficiency by 18%","X-ARAPUCA for DUNE: 3.7% PDE, filters cost 18% light","Liquid argon test: X-ARAPUCA without filters beats baseline by 18%","DUNE photon detector: X-ARAPUCA efficiency 3.7%, filters worse"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000437,"raw_usage":{"total_tokens":2059,"prompt_tokens":747,"completion_tokens":1312,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":491,"completion_tokens_details":{"reasoning_tokens":1182}},"tokens_in":491,"tokens_out":1312,"duration_ms":9214,"temperature":1.0,"reasoning_tokens":1182,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T22:01:41.579299+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-measure the reference SiPM's PDE directly at 127 nm in liquid argon at the actual incidence angles of this setup (~75°) rather than using the liquid-xenon-based angle correction; if the resulting reference PDE differs from the assumed 12.1%, the X-ARAPUCA's absolute PDE values in Table 9 shift by that factor. Alternatively, run the reference-method comparison with a different calibrated reference device (e.g., a second SiPM model or a photodiode-based detector) and check that the XA PDE remains 3.7% at 4.5 V overvoltage.","supporting_citations":[],"review_version":1}