Response to cosmic muons of scintillator-SiPM assemblies measured at different temperatures

A. Stifutkin; M. Chadeeva

REVIEW 4 major objections 5 minor 15 references

Response to cosmic muons of scintillator-SiPM assemblies measured at different temperatures

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Scintillator tiles wrapped in reflective film respond to muons about 10% less at −30°C than at +30°C; bare tiles show no temperature dependence.

desk verdict Careful measurement with a real confound: the film attribution is provisional because the bare-tile control ran at a different overvoltage. read the letter →

arxiv 2506.08203 v1 pith:QUFEEQ5E submitted 2025-06-09 physics.ins-det hep-ex

M. Chadeeva , A. Stifutkin This is my paper

classification physics.ins-dethep-ex

keywords scintillatortilessiliconphotomultiplierreflectivefilmtemperaturedependencecosmicmuonslightyieldlow-temperaturecalorimetryminimumionisingparticleresponse

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that the temperature dependence of a scintillator tile's response to muons comes from the reflective film wrapping, not from the scintillator or photosensor alone. Measurements at five temperatures between +30°C and −30°C show that tiles covered with reflective film lose about 1.3% of their light yield per 10 K of cooling, roughly 10% over the full range, while bare tiles show no statistically significant slope. The result matters for highly segmented calorimeters that plan to operate at low temperatures to suppress photosensor noise, because a room-temperature calibration would otherwise misstate the signal scale in cold operation. The paper concludes that wrapped scintillator-photosensor assemblies need to be tested and calibrated at their actual operating temperature.

What carries the argument

The workhorse is a per-channel correction chain applied to cosmic-muon pulse-height spectra. Gain is extracted from single-photoelectron peaks in noise spectra, optical crosstalk is estimated from the same spectra using a Poisson model for the firing of neighbouring cells, and the relative photon detection efficiency is corrected with a literature parametrisation of efficiency versus overvoltage, evaluated with the measured channel-wise breakdown voltages. The corrected spectra are fitted with a convolution of a Landau energy-loss distribution and a Gaussian resolution term; the maximum of the fitted curve defines the response to a minimum-ionising particle. This chain lets the paper compare tile configurations across temperature while removing the photosensor and electronics effects, so the residual slope can be assigned to the tile's optical construction.

What would settle it

Repeat the bare-tile measurement at the same 2.5 V overvoltage used for the film-covered tiles; if bare tiles then also show a slope near 1.3% per 10 K, the film attribution fails. A second check would be to measure the reflectance of the reflective film alone between −30°C and +30°C: if the film's reflectance is flat while wrapped tiles still show the slope, the effect is an interface or light-collection property rather than film reflectance.

Watch

Extended reading notes

Core claim

The central result is a measured temperature effect in the optical tile, not in the photosensor. For machine-wrapped square tiles produced for a hadronic calorimeter prototype, for thick hand-finished tiles with slices of reflective film on the large faces, and for previously measured hand-wrapped tiles of the same general design, the average light yield from cosmic muons increases with temperature at about 0.1 photoelectrons per kelvin, i.e., about 1.3% per 10 K, so that going from +30°C to −30°C costs roughly 10% of the response. For bare thick tiles, whose side surfaces carry a diffuse reflector coating but whose large faces are unfilmed, linear fits give slopes consistent with zero. Because the measurements correct each channel's gain, optical crosstalk, and relative photon detection efficiency at every temperature point, the paper attributes the residual slope to a combined effect of the plastic scintillator and the reflective film, plausibly a degradation of film performance on cooling that is not covered by manufacturer specifications.

Load-bearing premise

The key assumption is that the corrections for photosensor gain, crosstalk, and photon detection efficiency work equally well at the different overvoltages used: bare tiles were measured near 5 V overvoltage, while film-covered tiles were measured near 2.5 V, so a residual overvoltage-dependent effect could masquerade as a film effect.

Editorial extensions

If this is right

A calorimeter calibrated with muons at room temperature and operated near −30°C would need a temperature-dependent response correction of about 6–7% for wrapped tiles, i.e., roughly 1.3% per 10 K, to keep energy measurements unbiased.
The similar slope across machine-wrapped tiles, film-slice tiles, and previously measured hand-wrapped tiles implies that the effect is generic to reflective-film-wrapped scintillator assemblies rather than a specific tile geometry.
For detectors operated with temperature gradients, tile-to-tile response becomes position-dependent at the percent level even if the photosensors themselves are corrected, so temperature uniformity across the calorimeter becomes part of the energy-scale calibration.
Low-temperature operation still improves the signal-to-noise ratio because photosensor dark rate drops roughly an order of magnitude per 40 K, but the noise benefit is partially offset by the roughly 10% signal loss over the +30°C to −30°C range.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural follow-up is to measure the same wrapped tiles with a fixed light source, such as an LED or a radioactive source, instead of muons; this would separate true scintillator light-yield changes from changes in light collection and film reflectivity.
If the reflective film is the active element, substituting a reflector measured to be temperature-stable and repeating the muon test would directly test the mechanism and could yield a temperature-insensitive tile design.
A 10% signal loss on cooling brings low-energy signals closer to the threshold for tile hits in highly granular calorimeters, so the practical impact may be largest near the minimum-ionising-particle detection threshold rather than for high-energy jets.
The reported slope for wrapped tiles suggests that repeated heating-cooling cycles could reveal hysteresis or drift if film properties change irreversibly; a dedicated multi-cycle measurement would quantify whether the effect is reversible.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Marina,

Quick take: the paper is a decent engineering measurement with one confound that needs fixing before the headline claim is solid. The new bit is the bare-tile control: they show that the same TEST tiles with no reflective film have no temperature dependence, while with film slices added they see about 1.3% per 10 K decrease. That comparison is what makes the paper worth reading.

What they do well: the data taking is careful. They measure gain, crosstalk, and relative PDE at each temperature, run two full heating-cooling cycles per configuration, and the uncertainties on the light yield are a few percent. The effect is consistent across three configurations: AHCAL wrapped tiles, TEST tiles with film slices, and their earlier HGCAL result. That reproducibility argues the effect is real for film-covered assemblies.

The soft spot is the overvoltage mismatch. Bare TEST tiles were run at about 5 V overvoltage; film-slice and AHCAL runs were at 2.5 V. The PDE correction uses a curve from ref [14] for 'similar' SensL SiPMs, not the actual devices. At 5 V the PDE is closer to saturation, so the correction is less sensitive to overvoltage errors; at 2.5 V the slope of PDE vs overvoltage is steeper, so any mismatch between the ref [14] curve and the real devices would leave a temperature-dependent residual. That residual could masquerade as a film effect. The authors could have closed this by measuring bare tiles at 2.5 V, or by validating the PDE curve on their own SiPMs. They did neither.

Is this fatal? No. The effect is large and consistent, and the bare tiles at 5 V are a useful data point. But the film attribution is provisional until the overvoltage confound is removed.

Bottom line: worth a serious referee, and the referee should insist on the missing control. I would not cite the film conclusion as established, but I would cite it as a measured temperature dependence for wrapped tiles if I were writing about HGCAL.

Best, [Your name]

Referee Report

4 major / 5 minor

Summary. The manuscript reports measurements of the response to cosmic muons of plastic-scintillator tiles read out by SiPMs, at five temperatures between -30°C and +30°C. Three configurations are studied: bare TEST tiles, TEST tiles with slices of 3M ESR reflective film added above and below, and CALICE AHCAL tiles machine-wrapped in ESR film. For each channel and temperature, the light yield is derived from Landau-Gaussian fits of muon pulse-height distributions and corrected for gain, optical crosstalk, and relative photon-detection efficiency. The authors report that bare tiles show no temperature dependence, while all configurations with reflective film show a response that increases with temperature by about 1.3% per 10 K, i.e. about 10% over the full 60 K range. This effect is attributed to an interplay between the scintillator and the reflective film, with implications for calorimeters operated at low temperature.

Significance. If the result is correct, it is practically important: highly granular scintillator-SiPM calorimeters such as HGCAL are operated near -35°C, and a roughly 10% temperature-dependent response of the active elements over a 60 K range would require dedicated corrections. The work has genuine strengths: the experimental procedures are careful, with channel-wise breakdown-voltage and overvoltage estimation, single-photoelectron gain extraction, crosstalk estimation from noise spectra, and explicit correction for the relative PDE using an external parametrisation. The repeated heating-cooling cycles and the use of identical SiPM positions when comparing bare and film-covered TEST tiles are good design choices. The main drawback is that the critical control comparison is not balanced: bare tiles were measured at about 5 V overvoltage, whereas film-covered runs were performed at 2.5 V. Combined with the use of a PDE curve for SiPMs 'similar' to, but not identical with, those installed, this leaves a plausible overvoltage-dependent systematic path to the claimed film effect.

major comments (4)

[Section 3, Fig. 5] The central comparison in Fig. 5 is not balanced in overvoltage. The bare TEST tiles were measured at about 5 V overvoltage, while the TEST tiles with 3M film slices and the AHCAL tiles were measured at 2.5 V overvoltage. Because the SiPM PDE correction is the main temperature-dependent correction applied in Eq. 2.5, any residual error in that correction will produce an apparent temperature slope that depends on the operating overvoltage. At 5 V the PDE is closer to saturation and the local slope d(PDE)/d(OV) is smaller; at 2.5 V the same residual is amplified. The missing measurement is a bare-tile run at 2.5 V, which would separate the film variable from the overvoltage variable. This is a load-bearing gap for the paper's main claim that the reflective film causes the observed 1.3%-per-10-K effect.
[Eq. 2.4 and Sec. 2.2.3] The PDE correction factor in Eq. 2.4 is computed from a parametrisation of the PDE(OV) curve in ref. [14], which the text itself describes as applying to SensL SiPMs 'similar to' those used in the current experiment, not to the identical MicroFC-30035-SMT-WP devices. If the true PDE-versus-overvoltage shape differs from that parametrisation, the correction in Eq. 2.5 leaves a temperature-dependent residual proportional to d(PDE)/d(OV) and to any overvoltage drift with temperature. The stated uncertainty of the PDE factor is 1.5-2.5%, which is of the same order as the claimed effect of about 1.3% per 10 K. The authors should either validate the PDE curve for the actual SiPMs or quantify the maximal residual slope that could be produced by plausible deviations from the assumed curve.
[Section 3, Figs. 5-6] The AHCAL tiles are only measured in their machine-wrapped configuration, so they cannot serve as an independent control for the film effect. The conclusion that 'for all tile combinations with reflective film' the same temperature slope appears rests on a comparison between bare TEST tiles at one overvoltage and film-covered TEST and AHCAL tiles at a different overvoltage. A demonstration that the bare-tile null result survives at 2.5 V would make the AHCAL data interpretable as a confirmation rather than as a second measurement made under the same potentially biased overvoltage condition.
[Sec. 2.2.4, Eq. 2.5] Equation 2.5 includes a term P0_ij that is not defined in the text. If it denotes a pedestal or noise contribution, this should be stated explicitly and its temperature dependence should be discussed, since it enters directly into the light-yield values used to extract the slopes in Figs. 5 and 6. At present the reader cannot check whether P0_ij is constant or itself temperature-dependent. This is a clarity issue, but in a paper whose conclusions are based on small slopes it deserves attention.

minor comments (5)

[Sec. 2.2.3] The sentence 'The uncertainty of this factor gives the noticeable contribution to overall uncertainty at the level of 1.5-2.5% s' appears truncated and should be completed, presumably as '1.5-2.5% systematic uncertainty'.
[Fig. 3] The axis labels and legends in Fig. 3 contain garbled text such as 'C oTEST tile 2, cycle 1, T = 0' and 'T = 00 p.e.'; these should be corrected to include units such as 'T = 0 °C'.
[Eq. 2.3] The definition of q in Eq. 2.3 is introduced after the equation is displayed and the meaning of the ranges in Eq. 2.2 could be clearer; a brief sentence describing L_min and L_max as the pedestal-range boundaries would help.
[References] Ref. [15] is an accepted-but-unpublished self-citation used to support the claim that the same slope is observed for HGCAL-like wrapped tiles. Since the present paper draws a quantitative comparison with it, the authors should provide a preprint number, a public version, or at least a more detailed description of the compared configurations.
[Figs. 5-6] The error bars show the error of the mean over two heating-cooling cycles. Showing the individual cycle points as well would allow the reader to assess cycle-to-cycle reproducibility, which is relevant because systematic shifts between cycles would not be visible in the current presentation.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: direct temperature-response measurement with external PDE correction; the only same-author citation is qualitative and non-load-bearing.

full rationale

The paper's central result is a direct measurement of muon response versus temperature, not a derivation in which an output is identified with an input by construction. The light-yield correction in Eq. (2.5) multiplies the measured most probable value by a gain extracted from single-photoelectron spectra, a crosstalk factor from noise spectra, and a relative PDE ratio taken from the external parametrisation of ref. [14]. None of these factors is fitted to the temperature-dependent response that is then reported; the reported slopes are free linear fits to the corrected data points. Ref. [15] is a self-citation by the same authors, but it appears only in the conclusion as a qualitative statement that the observed slope is similar to previously measured wrapped HGCAL tiles; it is not used to derive Eqs. (2.4) or (2.5), to set the overvoltage, or to exclude alternative explanations, so it is not load-bearing. The mismatch between the bare-tile runs at about 5 V overvoltage and the film/AHCAL runs at 2.5 V is a legitimate systematic concern: if the ref. [14] PDE curve does not exactly describe the installed SiPMs, an overvoltage-dependent residual could masquerade as a film effect. However, this is a question of measurement correctness and confounding variables, not of circularity, because the claimed film effect is not equal by construction to any fitted parameter or cited premise. No circular step satisfying the required standard was identified.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

No invented entities. Two hand-chosen experimental settings (gate length, overvoltage) affect the result. The analysis relies on a Poisson crosstalk model and an external PDE(overvoltage) curve from ref [14].

free parameters (2)

Integration gate length = 235 ns
Chosen from pulse shape at -30 C (Section 2.2). A temperature-dependent pulse shape could bias the integral, but no strong evidence is presented for such a bias.
Overvoltage setting = 5 V for bare tiles, 2.5 V for film-covered tiles
Not fitted, but chosen to avoid ADC overflow. The central bare-versus-film comparison is made at different overvoltage, so this choice affects the conclusion.

assumptions (4)

domain assumption Poisson model for fired-cell statistics in optical crosstalk estimation (eq. 2.1)
Used to derive crosstalk probability c; if the model is wrong, the light yield correction is biased.
domain assumption PDE(overvoltage) parametrization from ref [14] applies to the SensL MicroFC-30035 SiPMs used here
Eq. 2.4 uses ref [14] for 'similar' SiPMs; the uncertainty is quoted at 1.5-2.5%, but applicability is not independently verified.
domain assumption Temperature does not change trigger acceptance, muon flux, or geometry
Setup is fixed in a climate chamber, but no explicit monitoring of these potential temperature-dependent factors is described.
domain assumption Reflective film is presumed temperature-independent per manufacturer, so the observed effect is attributed to tile-film interrelation
Section 4 invokes manufacturer claims; the paper itself notes the effect is not supported by manufacturer information and is 'most likely' a material interrelation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Response to cosmic muons of scintillator-SiPM assemblies measured at different temperatures." pith.science (2026). https://pith.science/paper/QUFEEQ5E

@misc{pith2026250608203,
  author       = {Pith},
  title        = {Pith review of: Response to cosmic muons of scintillator-SiPM assemblies measured at different temperatures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QUFEEQ5E}},
  note         = {Machine review of arXiv:2506.08203}
}

read the original abstract

Highly segmented calorimeters represent a modern trend in experimental particle physics aimed at improving the energy resolution with the particle flow reconstruction. The widely used and cost-effective solution is the structure assembled from small scintillator elements readout by tiny photosensors. The improvement of signal to noise ratio can be achieved by operating at low temperatures. The paper presents studies of temperature dependence of response to muons for elements comprised of plastic scintillator tiles with different options of reflective coverage and read out by silicon photomultipliers. No temperature dependence of response for bare tiles has been detected. With reflective film, the reduction of response with decreasing temperature is observed. The measured effect amounts to 10% in the range from +30oC to -30oC.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

15 extracted references · 14 canonical work pages

[14]

Gundacker et al.,Experimental time resolution limits of modern SiPMs and TOF-PET detectors exploring different scintillators and Cherenkov emission,Phys

S. Gundacker et al.,Experimental time resolution limits of modern SiPMs and TOF-PET detectors exploring different scintillators and Cherenkov emission,Phys. Med. Biol.65(2020) P025001

work page 2020
[1]

J.C. Brient,Improving the jet reconstruction with the particle flow method: An introduction, in Proceedings of the 11th International Conference on Calorimetry in High-Energy Physics (CALOR 2004), 2004

work page 2004
[2]

Morgunov,Calorimetry Design with Energy-Flow Concept, in10th International Conference on Calorimetry in High Energy Physics (CALOR 2002), Pasadena, CA U.S.A., 2002

V.L. Morgunov,Calorimetry Design with Energy-Flow Concept, in10th International Conference on Calorimetry in High Energy Physics (CALOR 2002), Pasadena, CA U.S.A., 2002

work page 2002
[3]

Thomson,Particle Flow Calorimetry and the PandoraPFA Algorithm, Nucl

M.A. Thomson,Particle Flow Calorimetry and the PandoraPFA Algorithm, Nucl. Instrum. Meth. A 611 (2009) 25

work page 2009
[4]

Gao,CEPC and SppC Status – From the completion of CDR towards TDR,Int.J.Mod.Phys.A36 (2021) 2142005

J. Gao,CEPC and SppC Status – From the completion of CDR towards TDR,Int.J.Mod.Phys.A36 (2021) 2142005

work page 2021
[5]

Abada et al.,FCC Physics Opportunities : Future Circular Collider Conceptual Design Report Volume 1,Eur.Phys.J.C79(2019) 474

A. Abada et al.,FCC Physics Opportunities : Future Circular Collider Conceptual Design Report Volume 1,Eur.Phys.J.C79(2019) 474

work page 2019
[6]

Sefkow, A

F. Sefkow, A. White, K. Kawagoe, R. Pöschl and J. Repond,Experimental tests of particle flow calorimetry, Rev. Mod. Phys.88(2016) 015003

work page 2016
[7]

CALICEcollaboration, Design, construction and commissioning of a technological prototype of a highly granular SiPM-on-tile scintillator-steel hadronic calorimeter,JINST 18(2023) P11018 [2209.15327]

work page Pith review arXiv 2023

Show all 15 references

[8]

CMS collaboration, The CMS HGCAL detector for HL-LHC upgrade,PoSICHEP2018(2019) 086

2019
[9]

Garutti and Y

E. Garutti and Y. Musienko,Radiation damage of SiPMs,Nucl. Instrum. Meth. A926 (2019) 69 [1809.06361]

2019 arXiv
[10]

Peralta,Temperature dependence of plastic scintillators,Nucl

L. Peralta,Temperature dependence of plastic scintillators,Nucl. Instrum. Meth. A883 (2018) 20 [1709.06458]

2018 arXiv
[11]

Carturan, F

S.M. Carturan, F. Pino, C.L. Fontana, S. Moretto, E. Zanazzi, R. Raggio et al.,Temperature effects on light yield and pulse shape discrimination capability of siloxane based scintillators,Eur. Phys. J. C 80(2020) 1057

2020
[12]

Simon and C

F. Simon and C. Soldner,Uniformity Studies of Scintillator Tiles directly coupled to SiPMs for Imaging Calorimetry,Nucl. Instrum. Meth. A620 (2010) 196 [1001.4665]

2010 arXiv
[13]

Klanner,Characterisation of SiPMs,Nucl

R. Klanner,Characterisation of SiPMs,Nucl. Instrum. Meth. A926 (2019) 36 [1809.04346]

2019 arXiv
[15]

Chadeeva, V

M. Chadeeva, V. Rusinov and A. Stifutkin,Temperature dependence of response to muons for plastic scintillator tiles wrapped in reflective film,Accepted for publication in BLPI(2025) . – 10 –

2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [14]

Gundacker et al.,Experimental time resolution limits of modern SiPMs and TOF-PET detectors exploring different scintillators and Cherenkov emission,Phys

S. Gundacker et al.,Experimental time resolution limits of modern SiPMs and TOF-PET detectors exploring different scintillators and Cherenkov emission,Phys. Med. Biol.65(2020) P025001

work page 2020

[2] [1]

J.C. Brient,Improving the jet reconstruction with the particle flow method: An introduction, in Proceedings of the 11th International Conference on Calorimetry in High-Energy Physics (CALOR 2004), 2004

work page 2004

[3] [2]

Morgunov,Calorimetry Design with Energy-Flow Concept, in10th International Conference on Calorimetry in High Energy Physics (CALOR 2002), Pasadena, CA U.S.A., 2002

V.L. Morgunov,Calorimetry Design with Energy-Flow Concept, in10th International Conference on Calorimetry in High Energy Physics (CALOR 2002), Pasadena, CA U.S.A., 2002

work page 2002

[4] [3]

Thomson,Particle Flow Calorimetry and the PandoraPFA Algorithm, Nucl

M.A. Thomson,Particle Flow Calorimetry and the PandoraPFA Algorithm, Nucl. Instrum. Meth. A 611 (2009) 25

work page 2009

[5] [4]

Gao,CEPC and SppC Status – From the completion of CDR towards TDR,Int.J.Mod.Phys.A36 (2021) 2142005

J. Gao,CEPC and SppC Status – From the completion of CDR towards TDR,Int.J.Mod.Phys.A36 (2021) 2142005

work page 2021

[6] [5]

Abada et al.,FCC Physics Opportunities : Future Circular Collider Conceptual Design Report Volume 1,Eur.Phys.J.C79(2019) 474

A. Abada et al.,FCC Physics Opportunities : Future Circular Collider Conceptual Design Report Volume 1,Eur.Phys.J.C79(2019) 474

work page 2019

[7] [6]

Sefkow, A

F. Sefkow, A. White, K. Kawagoe, R. Pöschl and J. Repond,Experimental tests of particle flow calorimetry, Rev. Mod. Phys.88(2016) 015003

work page 2016

[8] [7]

CALICEcollaboration, Design, construction and commissioning of a technological prototype of a highly granular SiPM-on-tile scintillator-steel hadronic calorimeter,JINST 18(2023) P11018 [2209.15327]

work page Pith review arXiv 2023