REVIEW 3 major objections 4 minor 47 references
Flow-dependent tagging of $^{214}$Pb decays in the LZ dark matter detector
T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash
Pith's one-line read LZ tracks single radon-chain atoms through its liquid xenon to tag 63% of 214Pb decays — the detector's leading low-energy background — for a 9% exposure cost.
desk verdict Directly measured 63% 214Pb tagging efficiency holds up; the charged-ion transport model extrapolation to 81 minutes is the main soft spot, but the paper is honest and deserves serious review. 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 load-bearing objects are the flow map, the ion-drift model, and the evolving selection volume. The flow map is a vector field interpolated from neutral 222Rn–218Po pair displacements using a 10-cm Gaussian kernel over the 11 nearest sample vectors, integrated forward with an adaptive Runge-Kutta streamline integrator; the ion-drift model, Eq. (4), gives the instantaneous drift velocity V+Z = VL + (VI − VL)(1 − ΔT/λ)exp(−ΔT/λ), with VI = −0.55 mm/s, VL = −0.38 mm/s, λ = 115 s, fitted to the charged 222Rn–218Po band. The tag volume itself is a cylinder around each projected streamline whose radii grow with time as ΔXYs < g + h(1 − exp(−ΔT/κ)) and |ΔZs| < i + j(1 − exp(−ΔT/λ)), with paramet
What would settle it
Measure the drift of 214Pb+ and 214Bi+ directly: the semi-charged 218Po–214BiPo pairs (where one of the two intervening ions is neutral) form a population whose vertical displacement at late times (ΔT > 40 min) depends on the true drift parameters. If their late-time vertical deviations fall where the 218Po+-based extrapolation predicts, the charged-branch assumption holds; if they systematically deviate — or if the 75% minus 63% efficiency gap persists once species-specific drift is incorporated — the transport model, not the neutralization or branching assumptions, is the limiting piece.
Extended reading notes
Core claim
The central claim is that 214Pb beta decays can be individually anticipated and rejected, not just statistically subtracted. The demonstration rests on a transport model: 222Rn–218Po decay pairs provide displacement vectors for both neutral atoms (bulk flow) and singly charged ions (flow plus electric-field drift), and these vectors are interpolated into a three-dimensional flow field with streamlines projected forward in time. A charged-ion drift model, fitted to the 218Po+ population with initial drift velocity −0.55 mm/s relaxing to −0.38 mm/s on a 115-s time constant, extends the trajectories of charged progeny; selection windows around both neutral and charged streamlines, tuned on pure
Load-bearing premise
The tag's charged branch assumes that an ion-drift model fitted to 218Po+ ions over the first 15 minutes — with one species-independent neutralization time of 49 minutes — correctly predicts where 214Pb+ and 214Bi+ ions go for up to 81 minutes.
Editorial extensions
If this is right
- The leading low-energy electron-recoil background in LZ's WIMP search is reduced by 63% at 9.0% exposure cost; the voxelized version of the tag was already deployed in the 2024 WIMP search with ϵtot as a floating parameter in the background model.
- Tagged 214Pb events form the first low-energy ER calibration dataset taken concurrently with science data; its spread around the predicted beta-decay band matches the LZ-tuned response model (KS p-value 0.319) and was used to confirm the origin of reduced-charge-yield 124Xe double-electron-capture events.
- The same transport model pairs 218Po with its granddaughter 214Po (via 214BiPo events) out to 81 minutes, providing a pure, high-statistics validation channel for any extension of the tag.
- The 12-point gap between modeled (75%) and observed (63%) tag efficiency is attributed to extrapolating the 218Po+-only drift calibration to 214Pb+ and 214Bi+; the paper identifies backward-projected streamlines from 214BiPo events as a concrete upgrade that would also recover neutralized-ion paths.
- Separately from the tag, the paper shows the detector's mixing state itself is a background control: the Minimal Mixing state reduces 218Po activity in the 5.5-tonne fiducial volume by nearly 20% relative to the Low Mixing state used for science data.
Reading between the lines
- I read the method as transferable: any liquid xenon (or other liquid noble) detector with slow, stable circulation and radon low enough that decay pairs are unambiguous could build the same pair-based flow map; the High Mixing state described here, where flow speeds approach the 0.4 mm/s ion drift speed and the neutral/charged pair populations blur, is the natural failure boundary.
- A direct measurement of 214Pb+ and 214Bi+ drift velocities — for instance from the semi-charged 218Po–214BiPo population that falls between the neutral and charged bands — would test whether the 75-to-63% efficiency gap is an ion-drift model error; if it is, species-specific drift parameters should push the tag toward the modeled 75%.
- The concurrent-calibration property may end up mattering beyond WIMP searches: every tagged 214Pb decay is a known low-energy beta in science data, so any electron-recoil-band analysis could use the tag as a live detector-response check without interrupting data taking for source injections.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper presents a method for tagging 214Pb beta decays in the LZ liquid-xenon TPC by exploiting the low 222Rn activity and stable, slow liquid flow in the detector. Sequential alpha decays (222Rn–218Po) are paired to construct a flow map and a charged-ion drift model; the resulting transport model is used to define time-evolving volume selections around predicted 214Pb trajectories. Applying these volumes costs 9.0% of exposure and removes 63% ± 6%(stat) ± 7%(sys) of the 214Pb ground-state decays in the 5.5 t fiducial volume, as measured directly by re-fitting the 18–75 keVee spectrum after removing tagged events. The tag has already been used in LZ's 2024 WIMP search, and the paper also demonstrates a concurrent low-energy ER calibration sample from tagged 214Pb decays. The central uncertainty in the method is the extrapolation of the charged-ion drift model, fitted to 218Po+ data out to ~15 min, to 214Pb+ and 214Bi+ trajectories out to 81 min; the authors explicitly identify this as the likely cause of the gap between the modeled (75%) and observed (63%) tagging efficiency.
Significance. If the result holds, this is an important advance for rare-event searches in liquid xenon: 214Pb is the leading low-energy ER background in LZ, and a demonstrated 63% tagging efficiency at 9% exposure cost is a significant background-mitigation tool. The paper's strongest point is that the headline efficiency is a direct spectral measurement (Section V.B.1), not a prediction from the transport model, so the central number does not collapse even if some model details are imperfect. The paper is also valuable for its detailed characterization of flow states and ion-neutralization times in a large LXe detector, and for introducing a concurrent calibration side band that has already been used in a published WIMP search. However, the transport-model extrapolation is a genuine limitation that affects the interpretation, portability, and the completeness of the quoted systematic uncertainty.
major comments (3)
- [Section V.C; Eq. (4), Eq. (7)] The charged-branch tag windows are built from an ion drift model, Eq. (4), fitted to 222Rn–218Po pairs with ΔT ≤ 15 min, then applied to 214Pb+/214Bi+ trajectories out to 64–81 min. The authors state in Section V.C that this is the likely source of the 75% vs 63% gap between modeled and observed tagging efficiency. Because the tag volumes themselves are defined by this model, systematic errors in the charged branch affect the composition of the tagged sample and the portability of the quoted efficiency to other flow states or drift fields. Yet the ±7% systematic on ϵtot in Section V.B.1 includes only the 214Pb-content prior, not the transport-model extrapolation. I would like a quantitative robustness test: vary VI, VL, and λ in Eq. (4) within the ranges allowed by the 218Po+ data and the 218Po–214BiPo pair distributions, and propagate the resulting changes to ϵtot. Absent this, the clai
- [Appendix A; Table III] The charge branching fractions and neutralization time τn used to interpret the tag efficiency are measured from the same neutral search windows that define the tag. This self-calibration loop, combined with the assumptions that progeny charge state is independent of progenitor charge and that all charged progeny neutralize with the same mean τn, means that the modeled efficiency of 75% is not an independent validation of the transport model. In particular, the 218Po–214Pb charge branching fraction of 0.48 ± 0.12 is derived with the same assumed τn and is then replaced by the 222Rn–218Po value in the model, so the quoted ±0.12 is not propagated. The paper is transparent about these assumptions, but the systematic impact of this circularity should be assessed—for example, by re-deriving τn and the branching fractions with an independent dataset (e.g., 218Po–214BiPo pairs at late ΔT) or by
- [Section V.B.1; Figure 12] The spectral fit measures ϵtot and fexp simultaneously, but assumes all non-214Pb components are reduced by fexp uniformly. The paper notes that this is an approximation because background populations are not spatially uniform. The proxy-based fexp of 9.3% ± 0.6% is consistent with the direct integration, which is reassuring, but the systematic uncertainty on ϵtot does not include any contribution from this approximation. A short quantitative statement of how large the non-uniformity effect could be (e.g., by comparing the spectral-fit ϵtot with the shifted-search ϵpair in the same energy window) would strengthen the claim.
minor comments (4)
- [Table II] The caption is dense and the phrase 'complimentary ϵ' appears to be a typo for 'complementary.' Also, the columns for ground-state vs excited-state measurements could be clearer about which uncertainties are statistical and which are systematic.
- [Eq. (8)] The one-sided relation ϵtot ≳ ϵpair + (1−ϵpair)fexp is useful, but the direction of the approximation could be stated more explicitly in the text, since the subsequent estimates use '≲' and '≳' interchangeably.
- [References] Reference [38] contains a typo: 'Elesvier' should be 'Elsevier.'
- [Section IV.B] The parameters in Eqs. (1)–(3) are described as being set 'by hand'; a short explanation of the sensitivity of the final tagging efficiency to these choices would be helpful, even if the final tag windows are tuned separately.
Circularity Check
No significant circularity: the central 214Pb tagging efficiency is a direct spectral measurement, not a model output.
full rationale
The headline claim, ϵtot = 63%, is obtained by a spectral re-fit of the 18–75 keVee single-scatter spectrum before/after removing tagged events (Sec. V.B.1, Fig. 12). The tag volumes are constructed from an empirical flow map and ion-drift model built from 222Rn–218Po pairs (Sec. IV), independent of the 214Pb spectrum being fit; the transport model is validated against 218Po–214BiPo pairs and then checked against the measured 214Pb tag performance. The charge-branching and neutralization inputs (Table III, Appendix A) are measured from the same pair searches used in the consistency Monte-Carlo, so the modeled 75% efficiency is a self-consistency loop rather than an independent first-principles prediction; however, it is explicitly compared with, and found to exceed, the measured 63%, and it is not the source of the quoted efficiency. The systematic uncertainty on ϵtot is traced to the 214Pb branching-ratio prior from Ref. [10] (itself a published LZ background result, not a parameter of this paper's tag construction), not to the tag volumes. The only other self-referential element is the calibration side band, where the tagged sample is compared to the NEST ER model after selection with a NEST-based ER-band cut; this could mildly precondition that cross-check, but it does not bear on the tag-efficiency claim. Overall, no load-bearing step reduces to its own input by construction, and the paper explicitly flags the charged-branch extrapolation as the likely cause of the model-vs-data gap rather than concealing it.
Assumptions & free parameters
free parameters (7)
- Neutral tag window parameters (Eq. 5-6) =
g=15 mm, h=40 mm, κ=800 s, i=30 mm, j=30 mm, λ=800 s
- Charged tag window parameters (Eq. 7) =
g+=15 mm, h+=30 mm, κ+=144 s, i+=20 mm, j+=100 mm, λ+=600 s
- Ion drift model parameters (Eq. 3-4) =
VI=-0.55 mm/s, VL=-0.38 mm/s, λ=115 s
- Initial pairing bounds (Eq. 1-3) =
a=25 mm, b=0.14 mm/s, c=0.10 mm/s, d=0.10 mm/s, τ=30 s, e=0.11 mm/s, f=0.32 mm/s, ζ=80 s
- Charge neutralization time τn =
49 ± 4 min
- Charge branching fractions (Table III) =
0.49 ± 0.01 (222Rn-218Po), 0.48 ± 0.12 (218Po-214Pb), 0.74 ± 0.05 (214Pb-214Bi)
- 218Po charged fraction in Section IV.C Monte Carlo =
50.3%
assumptions (6)
- domain assumption 214Pb ground-state branching fraction is 12.7(9)% (from [13], carried through [10]).
- domain assumption The Low Mixing flow state is stationary and reproducible during science data taking; unstable periods are excluded.
- ad hoc to paper Progeny charge state is independent of progenitor charge state, and all charged progeny neutralize with the same mean time τn.
- ad hoc to paper The charged 214Pb+ to 214Bi+ system is treated as a single continuously existing ion in charged streamlines.
- domain assumption Alpha-decay identification in the fiducial volume is sufficiently efficient and pure that misidentification is negligible.
- domain assumption Only neutral and +1 charge states of progeny meaningfully affect the spatial populations.
Cite this review
Pith. "Pith review of Flow-dependent tagging of $^{214}$Pb decays in the LZ dark matter detector." pith.science (2026). https://pith.science/paper/BGCUCMGV
@misc{pith2026250819117,
author = {Pith},
title = {Pith review of: Flow-dependent tagging of $^214$Pb decays in the LZ dark matter detector},
year = {2026},
howpublished = {\url{https://pith.science/paper/BGCUCMGV}},
note = {Machine review of arXiv:2508.19117}
}
abstract
The LUX-ZEPLIN (LZ) experiment is searching for dark matter interactions in a liquid xenon time projection chamber (LXe-TPC). This article demonstrates how control of the flow state in the LXe-TPC enables the identification of pairs of sequential alpha-decays, which are used to map fluid flow and ion drift in the liquid target. The resulting transport model is used to tag $^{214}$Pb beta-decays, a leading background to dark matter signals in LZ. Temporally evolving volume selections, at a cost of 9.0% of exposure, target the decay of each $^{214}$Pb atom up to 81 minutes after production, resulting in (63 $\pm$ 6$_{\mathrm{stat}}$ $\pm$ 7$_{\mathrm{sys}}$)% identification of $^{214}$Pb decays to ground state. We also demonstrate how flow-based tagging techniques enable a novel calibration side band that is concurrent with science data.
Figures
Figures from the paper (14 more)
Reference graph
Works this paper leans on
-
[1]
Low-energy spectral analysis of ϵtot Figure 12 shows the effect of the 214Pb tag on the 214Pb-dominated sample. The energy spectrum in this region includes the roughly flat, continuous electron- recoil background described in [10] (a combination of 214Pb, 85Kr, 136Xe, 212Pb, and solar neutrino-electron scattering), plus two peaks at 40.4 and 67.3 keV from...
-
[2]
Broad-spectrum analysis of ϵpair To simplify the application of the 214Pb tag to larger samples, such as the full single-scatter spectrum, the LXe-TPC is divided into 1.5-cm cubic voxels. Each voxel is flagged as “tagged” whenever it overlaps with any cylindrical exclusion region defined by Eq. (5–7), re- sulting in a voxelized tag volume that fully cover...
work page 2024
-
[3]
V. N. Lebedenko et al., Phys. Rev. D 80, 052010 (2009), arXiv:0812.1150 [astro-ph]
work page Pith review arXiv 2009
-
[4]
D. S. Akerib et al. (LUX), Phys. Rev. D 97, 102008 (2018), arXiv:1712.05696 [physics.ins-det]
work page Pith review arXiv 2018
-
[5]
J. B. Albert et al. (EXO-200), Phys. Rev. C 89, 015502 (2014), arXiv:1306.6106 [nucl-ex]
work page Pith review arXiv 2014
-
[6]
Physics reach of the XENON1T dark matter experiment
E. Aprile et al. (XENON), JCAP 04, 027 (2016), arXiv:1512.07501 [physics.ins-det]
work page Pith review arXiv 2016
- [7]
-
[8]
B. J. Mount et al. (LZ), (2017), arXiv:1703.09144 [physics.ins-det]
arXiv 2017
Show all 47 references
-
[9]
D. S. Akerib et al. (LZ), Nucl. Instrum. Meth. A 953, 163047 (2020), arXiv:1910.09124 [physics.ins-det]
2020 arXiv
-
[10]
Linehan et al
R. Linehan et al. , Nucl. Instrum. Meth. A 1031, 165955 (2022), arXiv:2106.06622 [physics.ins-det]
2022 arXiv
-
[11]
Aalbers et al
J. Aalbers et al. (LZ), Phys. Rev. Lett. 131, 041002 (2023), arXiv:2207.03764 [hep-ex]
2023 arXiv
-
[12]
Aalbers et al
J. Aalbers et al. (LZ), Phys. Rev. D 108, 012010 (2023), arXiv:2211.17120 [hep-ex]
2023 arXiv
-
[13]
C. M. Baglin, Nucl. Data Sheets 113, 1871 (2012)
2012
-
[14]
Singh et al
B. Singh et al. , Nucl. Data Sheets 160, 405 (2019)
2019
-
[15]
Zhu and E
S. Zhu and E. A. McCutchan, Nucl. Data Sheets 175, 1 (2021)
2021
-
[16]
J. B. Albert et al. (EXO-200), Phys. Rev. C 95, 025502 (2017), arXiv:1609.04467 [physics.ins-det]
2017 arXiv
-
[17]
B´ eet al., Table of radionuclides (Vol
M.-M. B´ eet al., Table of radionuclides (Vol. 4 - A = 133 to 252) (2008)
2008
-
[18]
D. S. Akerib et al. (LZ), Eur. Phys. J. C 80, 1044 (2020), [Erratum: Eur.Phys.J.C 82, 221 (2022)], arXiv:2006.02506 [physics.ins-det]
2020 arXiv
-
[19]
Pushkin et al
K. Pushkin et al. , Nucl. Instrum. Meth. A 903, 267 (2018), arXiv:1805.11306 [physics.ins-det]
2018 arXiv
-
[20]
Aprile et al
E. Aprile et al. (XENON), Eur. Phys. J. C82, 599 (2022), arXiv:2112.05629 [physics.ins-det]
2022 arXiv
-
[21]
Murra, D
M. Murra, D. Schulte, C. Huhmann, and C. Weinheimer, Eur. Phys. J. C 82, 1104 (2022), arXiv:2205.11492 [physics.ins-det]
2022 arXiv
-
[22]
Qian et al
Z. Qian et al. (PandaX-4T), JHEP 06, 147 (2022), arXiv:2112.02892 [physics.ins-det]
2022 arXiv
-
[23]
Cui et al., JINST 19, P07010 (2024), arXiv:2401.01638 [physics.ins-det]
X. Cui et al., JINST 19, P07010 (2024), arXiv:2401.01638 [physics.ins-det]
2024 arXiv
-
[24]
N. I. Chott and R. W. Schnee (LZ), AIP Conf. Proc. 2908, 080004 (2023), arXiv:2211.11857 [physics.ins-det]
2023 arXiv
- [25]
-
[26]
Aprile et al
E. Aprile et al. (XENON), Phys. Rev. D 110, 012011 (2024), arXiv:2403.14878 [hep-ex]. 21
2024 arXiv
-
[27]
Aalbers et al
J. Aalbers et al. (LZ), Phys. Rev. Lett. 135, 011802 (2025), arXiv:2410.17036 [hep-ex]
2025 arXiv
-
[28]
Aalbers et al
J. Aalbers et al. (LZ), JINST 19, P08027 (2024), arXiv:2406.12874 [physics.ins-det]
2024
-
[29]
B´ eet al., Table of radionuclides (Vol
M.-M. B´ eet al., Table of radionuclides (Vol. 2 - A = 151 to 242) , edited by B. I. des Poids et Mesures, Table of radionuclides, Vol. Vol.2 - A = 151 to 242 (2004)
2004
-
[30]
J¨ org, D
F. J¨ org, D. Cichon, G. Eurin, L. H¨ otzsch, T. Undagoitia Marrod´ an, and N. Rupp, Eur. Phys. J. C 82, 361 (2022), arXiv:2109.13735 [physics.ins-det]
2022 arXiv
-
[31]
Bradley, LUX Thermosyphon Cryogenics and Radon- related Backgrounds for the First WIMP Result , Ph.D
A. Bradley, LUX Thermosyphon Cryogenics and Radon- related Backgrounds for the First WIMP Result , Ph.D. thesis, Case Western Reserve University (May 2014)
2014
-
[32]
Rischbieter, Signal Yields and Detector Modeling in Xenon Time Projection Chambers, and Results of an Effective Field Theory Dark Matter Search Using LUX Data, Ph.D
G. Rischbieter, Signal Yields and Detector Modeling in Xenon Time Projection Chambers, and Results of an Effective Field Theory Dark Matter Search Using LUX Data, Ph.D. thesis, University of Albany (April 2022)
2022
-
[33]
Aprile, C
E. Aprile, C. E. Dahl, L. DeViveiros, R. Gaitskell, K. L. Giboni, J. Kwong, P. Majewski, K. Ni, T. Shutt, and M. Yamashita, Phys. Rev. Lett. 97, 081302 (2006), arXiv:astro-ph/0601552
2006 arXiv
-
[34]
D. S. Akerib et al. (LUX), JINST 13, P02001 (2018), arXiv:1710.02752 [physics.ins-det]
2018 arXiv
-
[35]
V. N. Solovov et al. , IEEE Trans. Nucl. Sci. 59, 3286 (2012), arXiv:1112.1481 [physics.ins-det]
2012 arXiv
-
[36]
Delaunay et al
B. Delaunay et al. , Izv. Akad. Nauk SSSR, Otdelenie Matematicheskii i Estestvennyka Nauk 7, 1 (1934)
1934
-
[37]
Perry, Background Mitigation in LZ and Next Gener- ation Rare Event Searches, Ph.D
E. Perry, Background Mitigation in LZ and Next Gener- ation Rare Event Searches, Ph.D. thesis, University Col- lege London (2024)
2024
-
[38]
J. B. Albert et al. (EXO-200), Phys. Rev. C 92, 045504 (2015), arXiv:1506.00317 [nucl-ex]
2015 arXiv
-
[39]
Aalbers et al
J. Aalbers et al. (LZ), Nucl. Instrum. Meth. A 1068, 169712 (2024), arXiv:2405.14732 [physics.ins-det]
2024 arXiv
-
[40]
Ahrens, B
J. Ahrens, B. Geveci, and C. Law, Visualization Hand- book (Elesvier, 2005) ISBN 978-0123875822
2005
-
[41]
Szydagis, N
M. Szydagis, N. Barry, K. Kazkaz, J. Mock, D. Stolp, M. Sweany, M. Tripathi, S. Uvarov, N. Walsh, and M. Woods, JINST 6, P10002 (2011), arXiv:1106.1613 [physics.ins-det]
2011 arXiv
-
[42]
Szydagis, A
M. Szydagis, A. Fyhrie, D. Thorngren, and M. Tripathi, JINST 8, C10003 (2013), arXiv:1307.6601 [physics.ins- det]
2013 arXiv
-
[43]
Aalbers et al
J. Aalbers et al. (LZ), J. Phys. G 52, 015103 (2025), arXiv:2408.17391 [nucl-ex]
2025 arXiv
-
[44]
Aalbers et al
J. Aalbers et al. (LZ), Phys. Rev. D 108, 072006 (2023), arXiv:2307.15753 [hep-ex]
2023 arXiv
-
[45]
Aalbers et al
J. Aalbers et al. (LZ), Phys. Rev. D (in press) (2025), arXiv:2503.05679 [hep-ex]
2025
-
[46]
P. J. W. Faulkner et al. (GridPP), J. Phys. G 32, N1 (2006)
2006
-
[47]
Britton et al
D. Britton et al. , Phil. Trans. Roy. Soc. Lond. A 367, 2447 (2009). Appendix A: Charge branching and ion neutralization The charge branching fractions for the 222Rn, 218Po, and 214Pb decays shown in Table III are determined from the pair excess in the neutral-progeny search w...
2009
Reviewed August 5, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.