REVIEW 3 major objections 3 minor 1 cited by
Neutrinoless double-beta decay search with the LEGEND experiment
T0 review · 3 major / 3 minor · reviewed 2026-08-10 · deepseek-v4-flash
Pith's one-line read The LEGEND-200 germanium detector, combined with the earlier GERDA and Majorana Demonstrator datasets, places a 90% confidence lower limit of $T_{1/2}>1.9\times10^{26}$ yr on neutrinoless double-beta decay in $^{76}$Ge.
desk verdict A new best 76Ge half-life limit from LEGEND-200, honestly presented but explicitly provisional because the paper's own background model is incomplete. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The detector concept is a self-triggering calorimeter: 140 kg of $^{76}$Ge-enriched germanium diodes act as both the source and the detector, immersed in 64 m$^3$ of liquid argon that serves as an active veto, inside a 590 m$^3$ water Cherenkov muon veto. Signal events deposit energy at a single site in one crystal; background rejection uses a multiplicity cut, pulse-shape discrimination against multi-site and surface events, and liquid-argon scintillation anti-coincidence. The statistical result is produced by a frequentist unbinned extended likelihood fit over the blinded $\pm25$ keV region around $Q_{\beta\beta}$, combining LEGEND-200 with the two predecessor experiments.
What would settle it
Look at the energies of the seven surviving events in the 1930–2190 keV window: the background-only model predicts a smooth distribution with no enhancement at $Q_{\beta\beta}=2039.061$ keV. If a corrected reanalysis, informed by the radio-assay campaigns of 2024, shows those seven events clustering at $Q_{\beta\beta}$, or if the resumed dataset produces an event rate in the window that rises faster than the background expectation, the background-only assumption behind the $1.9\times10^{26}$ yr limit is falsified.
Extended reading notes
Core claim
The central claim, stated as a preliminary result, is that a frequentist unbinned extended likelihood fit to the combined LEGEND-200 (48.3 kg·yr), GERDA, and Majorana Demonstrator exposures excludes $0\nu\beta\beta$ decay in $^{76}$Ge at half-lives below $T_{1/2}=1.9\times10^{26}$ yr at 90% confidence and reaches a median sensitivity of $2.8\times10^{26}$ yr. This is achieved in LEGEND-200 with a measured background index of about $5\times10^{-4}$ cnts/(keV·kg·yr), only a factor of about 2.5 above the design goal of $2\times10^{-4}$ cnts/(keV·kg·yr), and with only seven events surviving in the analysis window around $Q_{\beta\beta}=2039.061$ keV. On the same evidence base, the paper reports that the background was not fully understood: an excess appeared near $Q_{\beta\beta}$, and data taking stopped in February 2024 with most of the year spent identifying its source before a planned early-2025 restart. The future LEGEND-1000 phase is described as targeting a sensitivity beyond $10^{28}$ yr and an effective Majorana mass range of 9–21 meV, which is the band that would cover the inverted neutrino mass ordering.
Load-bearing premise
The reported half-life limit assumes that the seven events left in the analysis window are all known background and that the extra background discovered near the decay energy in 2024 adds no signal-like events; if that background model is incomplete, the limit would weaken.
Editorial extensions
If this is right
- If the limit is confirmed after the background anomaly is understood, germanium-76 searches will have crossed the $2\times10^{26}$ yr half-life threshold, and the next exposure at the design background index should reach the $10^{27}$ yr sensitivity target.
- The combined GERDA–Majorana–LEGEND analysis establishes a template for how phased experiments can pool blinded data sets to set a competitive limit before a final experiment reaches full exposure.
- LEGEND-1000, with a tonne of enriched germanium and a background goal of $10^{-5}$ cnts/(keV·kg·yr), is projected to cover $m_{\beta\beta}$ down to 9–21 meV and therefore to test the entire inverted ordering region in about a decade of running.
- The reported 2024 background excess means the current limit is preliminary; a reanalysis after the source is identified could either improve the background index toward the design value or, if some events are signal-like, change the limit.
Reading between the lines
- If the high background is traced to a specific nearby material, the resumed run offers a clean before/after test: the background index around $Q_{\beta\beta}$ should drop by the amount attributed to that component, which would validate the corrected model.
- The same analysis recipe—blinded window, topology cuts, argon veto, combined likelihood—could, in principle, be transferred to other $0\nu\beta\beta$ isotopes, but the dominant uncertainty would remain the completeness of each experiment's background model near its $Q$-value.
- A consequence the authors do not spell out: once the final background model is fixed, the half-life limit translates into a bound on the effective Majorana mass $m_{\beta\beta}$ whose size will be set by nuclear matrix element calculations; comparing that bound with neutrino oscillation and cosmological mass constraints would test the light-Majorana-exchange hypothesis.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This proceedings contribution (LIDINE 2024) describes the LEGEND program for searching neutrinoless double-beta decay in 76Ge, focusing on the LEGEND-200 first data-taking period (March 2023 to February 2024) and the planned LEGEND-1000 stage. The paper reports that a 48.3 kg·yr golden data set was analyzed, that 7 events survive all analysis cuts in the 1930–2190 keV window, that the background index is about 5×10^-4 cnts/(keV·kg·yr), and that a frequentist unbinned fit combining LEGEND-200 with Gerda and Majorana Demonstrator yields a 90% C.L. limit of T1/2 > 1.9×10^26 yr and a median sensitivity of 2.8×10^26 yr. The paper also discloses that data taking was interrupted in February 2024 because a higher-than-expected background appeared around the Q_beta_beta value, with much of 2024 devoted to understanding its origin.
Significance. If the quoted limit holds, it constitutes the strongest current lower bound on neutrinoless double-beta decay in 76Ge and strengthens the constraint on the effective Majorana neutrino mass. The manuscript is transparent: figures are marked 'Preliminary', the exposure and event count are stated, and the collaboration explicitly acknowledges the unexpected background component. The combination of three germanium experiments (Gerda, Majorana Demonstrator, and LEGEND-200) is a valuable step, and the statistical procedure (frequentist unbinned extended likelihood) is standard. The main significance hinges on the background model remaining valid after the unidentified component near Q_beta_beta is understood; the paper itself indicates this condition is not yet met, so the quoted central number is provisional.
major comments (3)
- [Abstract and Section 2] The central limit T1/2 > 1.9×10^26 yr and median sensitivity 2.8×10^26 yr are derived from the 7 events in the 1930–2190 keV window under a background model built from known sources. However, Section 2 states that data taking was interrupted in February 2024 because a 'higher than expected level of background events was discovered in the analysis of the data around the Q_beta_beta value' and that most of 2024 was spent understanding the origin of these events. If the unidentified component contributes inside the analysis window, the background expectation is underestimated and the fitted limit and sensitivity could change. Because the paper presents these numbers as the combined result without an explicit caveat in the abstract or conclusions, the headline claim is conditional on a background model that the collaboration itself treats as incomplete. The manuscript should state clearly that the quoted limit and sensitivity are preliminary and may be revised once the unexpected background component is modeled, or it should defer quoting a combined limit until the model is updated.
- [Section 2, right bottom plot] The background index is quoted as 'about 5×10^-4 cnts/(keV·kg·yr)' without any uncertainty. With only 7 surviving events in the analysis window, the Poisson statistical uncertainty is roughly 38%, and the possible contribution from the unexplained component near Q_beta_beta introduces an additional systematic uncertainty that is not quantified. The frequentist fit that produces the limit should either incorporate uncertainties on the background estimate or explicitly state how background systematics are treated; without this, the robustness of the 1.9×10^26 yr limit cannot be assessed.
- [Section 2, first paragraph] The selection of the golden data set of 48.3 kg·yr from the 76.2 kg·yr silver data set is not described. The paper only says that 'a subset ... was selected for the following 0νββ analysis' and that at present only BEGe, PPC, and ICPC detectors are used. Since the entire 0νββ result is based on this subset, the criteria for inclusion (e.g., detector performance, stability, data-quality cuts, any blinding procedure) should be stated so that the reader can judge whether the selection could introduce bias or affect the validity of the combined limit.
minor comments (3)
- [Abstract] The phrase 'one obtains a sensitivity on the half-life' should be 'median sensitivity', since the sensitivity quoted is a Monte Carlo expectation, not a measured value; this distinction is important for correct interpretation.
- [Section 2, Figure 3 caption] The captions of Figures 2 and 3 correctly label the plots 'Preliminary'. It would be helpful to add an explicit sentence in the text or captions that all numerical results in this paper (background index, limit, sensitivity) are preliminary and subject to revision after the current background investigation is completed.
- [Section 1] The conversion between the projected background index '0.6 cnts/(FWHM·t·yr)' and '2×10^-4 cnts/(keV·kg·yr)' appears to assume a FWHM of about 3 keV rather than the 2.5 keV FWHM mentioned in the Figure 1 caption; please clarify the conversion.
Circularity Check
No significant circularity: the reported limit is a direct experimental bound from combined exposures, not a quantity defined by its own fit parameters.
full rationale
The paper's central claim, a 90% C.L. limit T1/2 > 1.9x10^26 yr for neutrinoless double-beta decay in 76Ge, is obtained from a frequentist unbinned extended likelihood fit combining the exposures of GERDA, the Majorana Demonstrator, and LEGEND-200. This is a direct experimental measurement from external datasets, not a derived quantity that has been engineered to equal its inputs. The quoted median sensitivity of 2.8x10^26 yr is a Monte Carlo expectation under the background hypothesis, so it is not a fitted parameter subsequently relabeled as a prediction. The 2νββ spectral normalization uses the half-life measured by GERDA (Ref. [4]), but that is an independent input used only to model a background component; it is not the 0νββ half-life being claimed. The paper's own admission that a higher-than-expected background level was discovered near Qββ and that data taking was interrupted is a caveat about the completeness of the background model, and it may affect the reliability or final value of the limit, but it does not make the derivation circular. The LEGEND design-report self-citations (Refs. [3], [5]) describe the experimental program and future plans; they are not load-bearing for the numerical limit presented here. No equation in the paper defines the result in terms of itself, and no fitted parameter is renamed as a prediction. The appropriate finding is therefore no significant circularity.
Assumptions & free parameters
assumptions (3)
- domain assumption The 2nu betabeta background contribution is correctly normalized using the half-life measured by Gerda [4].
- standard math The frequentist unbinned extended likelihood fit and asymptotic formulas yield valid 90% C.L. limits for the small event count in the analysis window.
- domain assumption Pulse shape discrimination and LAr anti-coincidence cuts do not reject signal-like single-site events in the analysis window; the signal efficiency is known and stable.
Cite this review
Pith. "Pith review of Neutrinoless double-beta decay search with the LEGEND experiment." pith.science (2026). https://pith.science/paper/WO5FY67N
@misc{pith2026250110046,
author = {Pith},
title = {Pith review of: Neutrinoless double-beta decay search with the LEGEND experiment},
year = {2026},
howpublished = {\url{https://pith.science/paper/WO5FY67N}},
note = {Machine review of arXiv:2501.10046}
}
abstract
Neutrinoless double-beta decay is a nuclear decay, given as $(A,Z) \rightarrow (A, Z+2) +2e^{-}$, with deep consequences for the understanding of our universe. A strong experimental program is underway to search for this transition with many proposed experiments using different technologies. In this article the LEGEND experiment, which uses $^{76}$Ge as the isotope of interest, will be described. We will discuss both the first stage, LEGEND-200, which is now taking data at the Laboratori Nazionali del Gran Sasso of INFN in Italy, and the future stage, LEGEND-1000. LEGEND-200 has analyzed a first sample of data (48.3 kg$\cdot$yr) collected from March 2023 to February 2024 with a background index not far away from its goal of 2$\times$10$^{-4}$ cnts/(keV$\cdot$kg$\cdot$yr). Combining the LEGEND-200 data with those of \textsc{Gerda}\ and \textsc{Majorana Demonstrator}\ one obtains a sensitivity on the half-life of 0$\nu\beta\beta$ decay in $^{76}$Ge of $T_{1/2} > $ 2.8 $\times$ 10$^{26}$ yr at 90\% C.L. and a limit on $T_{1/2} > $ 1.9 $\times$ 10$^{26}$ yr at 90\% C.L.
Forward citations
Cited by 1 Pith paper
-
Nuclear Physics Confronts Relativistic Collisions Of Isobars
RHIC isobar data are explained by different shapes of 96Ru and 96Zr, with 96Zr showing a large octupole deformation, so nuclear structure uncertainty, not the magnetic field, dominates the observed ratios.
Reference graph
Works this paper leans on
-
[1]
(GerdaCollab.), Final Results of GERDA on the Search for Neutrinoless Double-𝛽 Decay, Phys
M.Agostini et al. (GerdaCollab.), Final Results of GERDA on the Search for Neutrinoless Double-𝛽 Decay, Phys. Rev. Lett. 125, 252502 (2020), [arXiv:2009.06079]
arXiv 2020
-
[2]
I. J. Arnquistet al. (MajoranaCollab.), Final Result of the Majorana Demonstrator’s Search for Neutrinoless Double-𝛽 Decay in 76Ge, Phys. Rev. Lett. 130, 062501 (2023), [arXiv:2207.07638v1]
arXiv 2023
-
[3]
N. Abgrall et al. (LEGEND Collab.),The Large Enriched Germanium Experiment for Neutrinoless Double Beta Decay (LEGEND), AIP Conf. Proc. 1894, 020027 (2017) [arXiv:1709.01980v1]
arXiv 2017
-
[4]
M. Agostini et al. (GerdaCollab.), Final Results of GERDA on the Two-Neutrino Double-𝛽 Decay Half-Life of 76Ge, Phys. Rev. Lett. 131, 142501 (2023), [arXiv:2308.09795]
arXiv 2023
-
[5]
N. Abgrall et al. (LEGEND Collab.),LEGEND-1000 Preconceptual Design Report, [arXiv:2107.11462v1]. – 4 –
Reviewed August 10, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.