REVIEW 3 major objections 4 minor 10 references
Characterization of Two PMT Models for the IceCube Upgrade mDOM
T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read The paper establishes that two candidate photomultiplier tubes for the IceCube Upgrade mDOM—the Hamamatsu R12199-01 MOD HA and the HZC XP82B2F—meet their manufacturer specifications on gain, timing, quantum efficiency, and dark noise…
desk verdict A useful, clearly scoped PMT comparison for the IceCube mDOM, but the HZC 'excellent noise' claim leans on three tubes and no error bars in the key tables. 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 procedure is the charge-distribution analysis: time-integrated PMT pulses are histogrammed, and a model fit separates the pedestal, single-photoelectron, and multi-photoelectron peaks so that gain and peak-to-valley ratio are extracted. Timing is measured with transit-time histograms whose Gaussian width defines the transit-time spread, and early and delayed pulse fractions are counted relative to the main peak. Quantum efficiency is measured with monochromator-selected light against a reference photodiode. For dark noise, inter-pulse time differences are binned to separate a fast correlated component, attributed to radioactivity and glass scintillation, from an uncorrelated thermionic component, and the whole measurement is repeated in a cold freezer down to -20 °C.
What would settle it
Measure dark noise rates at -20 °C on a production-size sample of HZC XP82B2F tubes, say 30 or more, operated at the same gain as the Hamamatsu tubes; if the mean uncorrelated dark noise rate exceeds the paper's benchmark or varies widely across tubes, the conclusion that the HZC model has excellent low-temperature noise behaviour would be undermined.
Extended reading notes
Core claim
The core claim is that the two PMT models show good agreement with manufacturer specifications across the measured quantities. At a nominal gain of 5 × $10^{6}$, the Hamamatsu tubes required a mean high voltage of 1148 V and the HZC tubes 1205 V; the Hamamatsu PMTs had a mean transit time spread of 1.49 ns versus 1.91 ns for the HZC PMTs, while quantum efficiency at 390 nm was nearly equal at about 26.8% and 27.0%. The temperature-dependent tests showed that at -20 °C the total dark noise rate averaged 35 $s^{-1}$ for the Hamamatsu and 67 $s^{-1}$ for the HZC, with uncorrelated rates of 12 $s^{-1}$ and 22 $s^{-1}$, respectively. Both models therefore show low-temperature noise behaviour well below the 100 $s^{-1}$ level and, according to the paper, qualify for use in future IceCube optical modules.
Load-bearing premise
The cold-temperature noise conclusion rests on only 12 Hamamatsu and 3 HZC tubes, and one of those three HZC tubes strongly pulls up the room-temperature average, so the claim of excellent HZC noise behavior holds only if those few tubes represent the production model.
Editorial extensions
If this is right
- Both vendors meet the mDOM requirement on gain, timing, quantum efficiency, and cold dark noise, so the module design is not held hostage to a single PMT supplier.
- Dark noise rates below 100 s^-1 at -20 °C mean background from PMT noise should not dominate the trigger or hit-selection logic in the deep-ice environment.
- The Hamamatsu PMT's lower transit time spread and lower room-temperature dark noise are balanced by the HZC PMT's larger photocathode area, so the final choice depends on system-level trade-offs such as effective photosensitive volume and operating gain.
- The independently measured parameters can be used directly as quality-assurance benchmarks when the mDOM production batches are tested.
- The temperature-dependent gain drift identified here establishes that operating voltage will need adjustment as a function of temperature to keep gain fixed in the deployed modules.
Reading between the lines
- Editorial inference: if the three HZC tubes tested at low temperature are representative, then the larger HZC photocathode could give the mDOM a modestly larger effective sensitive area at the cost of a slightly wider timing spread and higher dark noise; a full optical-module simulation would be needed to see which effect wins.
- Editorial inference: the paper's cold-noise conclusion would be on much firmer ground if repeated on a production-size HZC sample; a natural extension is to measure a few dozen tubes across several manufacturing batches at -20 °C.
- Editorial inference: the correlated dark-noise component attributed to glass scintillation suggests that thermal cycling or different glass batches could change the low-temperature noise floor, so aging and cycling tests would be a useful next step beyond the static cold measurements reported here.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This ICRC proceedings paper reports characterization measurements of two photomultiplier tube models considered for the IceCube Upgrade multi-PMT optical module (mDOM): the Hamamatsu R12199-01 MOD HA and the HZC XP82B2F. The authors directly measure gain, timing properties, quantum efficiency, and dark noise rate as a function of temperature using standard laboratory equipment, including a pulsed LED, oscilloscope, monochromator, and a temperature-controlled dark freezer. Room-temperature results over 100 Hamamatsu and 45 HZC tubes are summarized in Table 1, and low-temperature dark noise rates for 12 Hamamatsu and 3 HZC tubes are given in Table 2. The paper concludes that both PMT models agree with manufacturer specifications and show excellent noise behavior at the -20 C temperatures relevant for the IceCube Upgrade.
Significance. The measurements are directly relevant to PMT selection for the IceCube Upgrade mDOM, and the companion papers [5,6] provide detailed documentation of the two test programs. Strengths include the use of standard, adequately specified measurement methods; direct calibration of gain from single-photoelectron charge distributions; direct measurements of timing, quantum efficiency, and dark noise without circular reuse of fitted values; and large sample sizes for the room-temperature properties. If the low-temperature conclusions hold, the paper provides a concise comparative record supporting both tubes as viable mDOM candidates. The main limitation is that the HZC dark-noise conclusion rests on only three tubes, with one tube explicitly noted as a strong room-temperature outlier, and no per-tube low-temperature values or uncertainties are reported.
major comments (3)
- [Sec. 4.1 / Table 2] The central conclusion that "both PMT models ... show excellent noise behaviour at the low-temperatures relevant for possible use in future IceCube optical modules" is not fully supported for the HZC model. The HZC low-temperature averages (67 s-1 total, 22 s-1 uncorrelated) are based on three tubes, no per-tube spread or uncertainty is given, and the paper itself states in Sec. 4.1 that one of these three tubes had particularly high room-temperature dark noise and strongly biased the averages. Because the same small subset is used for the -20 C benchmark, a single non-representative tube could dominate the low-temperature averages as well. Per-tube values at -20 C, or additional HZC tubes, are needed before "excellent noise behaviour" can be claimed for the HZC model as a whole.
- [Sec. 4.1 / Table 2] No uncertainties or sample spreads are reported for any entry in Table 2. Since these dark noise rates are the load-bearing evidence for the low-temperature conclusion, reporting only single averaged numbers makes it impossible to judge whether the differences between models, or even the claim that rates are "well under 100 s-1", are robust. Standard deviations, standard errors, or per-tube values should be added.
- [Sec. 3 / Table 1 / Sec. 5] The conclusion that the results are "in good agreement with the specifications as provided by the manufacturers" cannot be checked from the data as presented, because Table 1 reports only mean values and min-max ranges across tubes, with no measurement uncertainties. The min-max ranges give the spread across samples, but not the precision of each individual measurement; the relevant manufacturer specifications and the uncertainty on the reported means should be stated explicitly.
minor comments (4)
- [Sec. 4] The first paragraph of Sec. 4 says twelve Hamamatsu PMTs were tested at low temperatures, but the next sentence refers to "the low-temperature tests on the two Hamamatsu PMTs"; this should be corrected to twelve or the discrepancy explained.
- [Fig. 4] The vertical axis labels in Figure 4 appear as "PMT s" on both panels; these should be replaced with "Number of PMTs" or similar.
- [Eq. (3.1) and Fig. 3 (right)] The power-law fit parameters a and b in the gain calibration are not defined; please define b explicitly as the slope and state how its uncertainty (e.g., 6.6 +/- 0.1 in Fig. 3) was obtained.
- [Table 2] The note that the HZC PMTs were operated at a gain of 1e7 while the Hamamatsu PMTs were at 5e6 is important for interpreting the dark-noise comparison; ideally the HZC tubes would also be measured at 5e6, or the model-level comparison should be explicitly limited to the two operating conditions used.
Circularity Check
No circularity: characterization results are direct measurements, and the cited companion papers provide supporting detail rather than load-bearing assumptions.
full rationale
The paper is an experimental PMT characterization and contains no derivation chain that re-inserts fitted values or predicts quantities from its own inputs. Gain (Eq. 3.1) is computed directly from fitted SPE and pedestal peak positions; TTS, early/delayed pulse fractions, QE, and dark noise rates are all measured quantities reported with histograms and tables. The conclusion that both PMT models agree with manufacturer specifications and show excellent low-temperature noise behavior is an external consistency claim, not an internally derived prediction. The self-references to [5] and [6] are explicitly presented as sources of more detailed characterization results ('Parts of these proceedings are a condensed version of and a comparison between the PMT characterization results presented for the Hamamatsu PMT in [5] and the HZC PMT in [6]'), and the central data in Tables 1 and 2 are reported in this proceedings itself. The paper honestly flags the small HZC sample and the biasing effect of one high-dark-noise tube on the room-temperature averages; this is a statistical representativeness limitation, not a circularity. No quantity is defined in terms of the conclusion, no fitted parameter is renamed as a prediction, and no uniqueness theorem or ansatz is imported from the authors' prior work to force the result.
Assumptions & free parameters
free parameters (1)
- Gain calibration power-law parameters (a, b) =
Per-PMT; example HZC tube has b=6.6±0.1 and reaches 5×10^6 gain at 1282 V
assumptions (4)
- domain assumption Charge distribution is a pedestal plus Gaussian n-photoelectron peaks, modeled as in [7].
- domain assumption Dark noise at room temperature is dominated by thermionic photocathode emission and decreases with cooling.
- domain assumption Correlated dark noise at short hit-time differences arises from radioactive decays and scintillation in the PMT glass.
- domain assumption The small test samples are representative of the production PMT populations.
Cite this review
Pith. "Pith review of Characterization of Two PMT Models for the IceCube Upgrade mDOM." pith.science (2026). https://pith.science/paper/EPTK3YYJ
@misc{pith2026190808446,
author = {Pith},
title = {Pith review of: Characterization of Two PMT Models for the IceCube Upgrade mDOM},
year = {2026},
howpublished = {\url{https://pith.science/paper/EPTK3YYJ}},
note = {Machine review of arXiv:1908.08446}
}
read the original abstract
The IceCube Upgrade will expand the IceCube Neutrino Observatory with nearly 800 new optical modules. A large fraction of these will be multi-PMT optical modules (mDOMs), featuring 24 PMTs pointing uniformly in all directions, providing an almost homogeneous angular coverage and providing an effective photosensitive area more than twice that of current IceCube optical modules. Two PMT models from different manufacturers are currently considered for use in the mDOM: a 3.5 inch PMT from HZC Photonics and a 3 inch PMT from Hamamatsu. Both PMTs have been characterized in terms of gain, timing, quantum efficiency and dark noise rate as a function of temperature. The obtained characterization results are presented here.
Figures
Figures from the paper (4 more)
Reference graph
Works this paper leans on
- [5]
-
[6]
M. A. Unland Elorrieta et al., Journal of Instrumentation 14 (2019) P03015
work page 2019
-
[1]
write newline
" write newline "" before.all 'output.state := FUNCTION blank.sep after.quote 'output.state := FUNCTION fin.entry output.state after.quoted.block = 'skip 'add.period if write newline FUNCTION new.block output.state before.all = 'skip output.state after.quote = after.quoted.block 'output.state := after.block 'output.state := if if FUNCTION new.sentence out...
-
[2]
IceCube Collaboration, M. G. Aartsen et al., JINST 12 (2017) P03012
2017
-
[3]
IceCube Gen2 Collaboration, M. G. Aartsen et al., http://arxiv.org/abs/1412.5106 arXiv:1412.5106
-
[4]
Adri\' a n-Mart\' i nez et al., J.\,Phys
KM3NeT Collaboration, S. Adri\' a n-Mart\' i nez et al., J.\,Phys. G43 (2016) 084001
work page 2016
-
[7]
van Eijk et al., Journal of Instrumentation 14 (2019) P07009
D. van Eijk et al., Journal of Instrumentation 14 (2019) P07009
work page 2019
-
[8]
E. Bellamy et al., Nucl. Instrum. Methods Phys. Res. A 339 (1994) 468 -- 476
work page 1994
Show all 10 references
-
[9]
Lubsandorzhiev et al., Nucl
B. Lubsandorzhiev et al., Nucl. Instrum. Methods Phys. Res. A 442 (2000) 452 -- 458
2000
-
[10]
a t Erlangen-N \
L. Classen, PhD thesis Friedrich-Alexander-Universit \"a t Erlangen-N \"u rnberg (FAU) (2001)
2001
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.