Pith. sign in

REVIEW 3 major objections 4 minor 10 references

Self-Contained, Cooled SiPM Array for Scintillation Spectroscopy

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read A thermoelectrically cooled, fan-vented silicon photomultiplier array, held at -20 °C, cuts dark count by about 1000-fold and delivers gamma-ray spectra comparable to a photomultiplier tube above 25 keV in NaI.

desk verdict Useful engineering data on a cooled SiPM array, but the 'nearly identical to PMT' claim has no PMT metrics behind it, and the portability framing outruns the actual vacuum-pump/AC-fan apparatus. read the letter →

arxiv 1908.08852 v1 pith:NSZGHS3J submitted 2019-08-23 physics.ins-det

A. Ponento , C. J. Martoff , D. Jones , E. Kaczanowicz This is my paper · ORCID
classification physics.ins-det
keywords siliconphotomultipliergamma-rayspectroscopythermoelectriccoolingscintillationdetectordarkcountrateNaI(Tl)portablegammathermo-stabilization
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 demonstrates that a large-area silicon photomultiplier (SiPM) array becomes a practical gamma-ray spectrometer when cooled to -20 °C with a compact, fan-vented thermoelectric system, no water line required. At that temperature the dark count rate in an air-coupled 2-inch NaI scintillator setup is reduced by roughly a factor of 1000 for energies above 40 keV, and the cooled array resolves the same Cs-137 and Ba-133 features as a photomultiplier tube for energies above about 25 keV. The point of the demonstration is portability: a rugged, self-contained SiPM spectrometer of this kind could replace vacuum phototubes in field applications such as cargo inspection, where the PMT's fragility and sensitivity to magnetic fields are drawbacks.

What carries the argument

The mechanism that carries the argument is the thermal suppression of thermally generated carriers in silicon: SiPM dark count falls by roughly an order of magnitude for every 30 K of cooling, so reaching -20 °C from room temperature removes most of the noise that otherwise buries few-photon pulses. The engineering realization is a three-stage heat-removal chain: the SiPM is pressed against a sapphire window inside a rough vacuum and thermally coupled to an inner thermoelectric cooler; four parallel thermoelectric coolers outside the vacuum pull heat from the inner cooler's hot side; and a large aluminum heat sink with two AC fans vents the total 170 W of waste heat to room air. This chain makes the -20 °C operating point reachable without water cooling, which is what converts the detector from a bench device into a self-contained portable instrument.

What would settle it

Install an independently calibrated temperature sensor directly on the SiPM package and repeat the dark-count and Cs-137/Ba-133 measurements while logging that temperature; if the true silicon temperature is significantly warmer than -20 °C, or if the measured dark-count drop from room temperature to -20 °C does not follow the roughly order-of-magnitude-per-30-K trend, then the cooling explanation for the improved spectra is not supported.

Watch

Extended reading notes

Core claim

The central claim is that active cooling is sufficient, not just helpful, for SiPM-based scintillation spectroscopy. With the array at -20 °C, the dark count measured through the spectroscopy amplifier fell from 6630 kHz to 190 kHz at a 25 mV threshold and from 360 kHz to 0.01 kHz at a 125 mV threshold, so the highest-threshold background dropped by roughly four orders of magnitude. In pulse-height spectra, the Cs-137 photopeak resolution improved from 20% FWHM at room temperature to 16% when cold, and the cold spectrum showed the Compton edge, Compton plateau, and backscatter peak that were barely visible at room temperature; for Ba-133, the 356 and 303 keV peaks became distinguishable and the 81 and 31 keV peaks appeared. The authors conclude that, even with non-optimized air coupling to the NaI crystal, the cooled SiPM performs nearly identically to a PMT for energies above 25 keV in NaI.

Load-bearing premise

The load-bearing premise is that the board-reported -20 °C is the actual SiPM temperature during the spectra, and that the device is self-contained in practice — the prototype still needs an external roughing pump, so the portability claim assumes a sealed gas-filled enclosure can replace it without changing the thermal behavior.

Editorial extensions

If this is right

  • A fan-cooled, thermoelectrically cooled SiPM package can be made portable and self-contained, because the air-cooled version performed like the water-cooled version in the same setup.
  • Above roughly 25 keV in NaI, such a detector reproduces the photopeak, Compton edge, and backscatter features that nuclide identification relies on, despite the loss of low-energy sensitivity.
  • The remaining low-energy limit is set by dark noise rather than by the SiPM itself, so improving cooling and light collection should push the threshold to lower energies.
  • The design improvements the paper lists — a sealed gas-filled or getter-pumped enclosure, fewer thermal junctions, and multi-stage coolers — should reach -30 °C or below and widen the useful window.

Reading between the lines

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

  • If the same dark-count scaling holds below -20 °C, reaching -30 °C should reduce the working dark rate by roughly another factor of two, likely turning the 31 keV Ba-133 shoulder into a distinct peak and lowering the spectral floor below 25 keV.
  • The strong threshold dependence of the cooled dark count implies that the remaining noise is concentrated at small pulse heights; a pulse-shape or timing cut that rejects fast dark pulses might recover spectroscopy below 25 keV even before deeper cooling is engineered.
  • Because the cooled array already matches a PMT with an ungreased air gap, adding optical coupling compound should increase light collection and could make the SiPM's photopeak resolution better than the PMT's rather than equal to it.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper describes a thermoelectrically cooled silicon photomultiplier (SiPM) array enclosed in a rough vacuum chamber and coupled through a glass window to a 2"x2" NaI scintillator. The authors report that cooling the SiPM array to -20 C reduces its dark count rate by factors ranging from about 36 to 10^5 depending on discriminator threshold (Table 2), and they show pulse-height spectra for Cs-137 and Ba-133 at room temperature and at -20 C, alongside spectra taken with a photomultiplier tube (PMT) in the same setup. The central claim, stated in Section 5, is that with non-optimized optical coupling the cooled SiPM performance is 'nearly identical' to that of a PMT for energies above 25 keV in NaI, making the system adequate for field gamma-ray spectroscopy. The paper also describes the thermal design, including an air-cooled heat sink replacing an earlier water-cooled version, and discusses straightforward improvements for lower temperatures.

Significance. If the central claim is quantitatively established, the work is a useful instrumentation contribution: it demonstrates that a compact, TEC-cooled SiPM array can resolve gamma-ray photopeaks and low-energy features (including the 31 keV Ba-133 line as a shoulder) that are invisible at room temperature, and it provides a practical thermal design with estimated heat loads. The strengths of the paper are its direct experimental approach, the use of standard calibration sources with a linear energy scale, and the honest discussion of limitations in Section 5 (vacuum pump requirement, proof-of-principle mechanics). However, the paper's headline comparative claim against a PMT rests entirely on qualitative visual comparisons; no PMT resolution or valley-to-peak metrics are reported, and no uncertainties are given for any rate, resolution, or dark-count value. These are missing-support issues rather than internal inconsistencies, and they are addressable with additional analysis of the already-recorded data.

major comments (3)
  1. [Section 4.2, 4.3, and 5] The central conclusion that the cooled SiPM performance is 'nearly identical to that of a PMT for energies above 25 keV' is unsupported because no PMT performance metric is reported anywhere. Section 4.2 gives the SiPM photopeak FWHM (20% at 25 C, 16% at -20 C) and valley-to-peak ratio (18% to 9.6%), and Section 4.3 says the SiPM shows the same peaks as the PMT with 'slightly worse' valley-to-peak ratios, but the corresponding PMT FWHM and valley-to-peak values from the same air-coupled geometry are never given. Please add a table with the PMT FWHM and valley-to-peak values (and ideally the same quantities for both SiPM temperatures) so that 'nearly identical' can be quantitatively evaluated.
  2. [Section 4.1, Table 2] The dark-count reduction factors are presented without uncertainties or replicate measurements. Table 2 lists single dark-count values at each threshold and temperature, and the text quotes reductions such as 'a factor of about 1000' and 'factors ranging up to 100,000' without statistical errors. The temperature readout from the AiT interface board is also quoted as -20 C with no stated accuracy. Since the dark-count reduction is the physical basis for the spectroscopic improvement, please provide uncertainties (or at least multiple runs) and state the accuracy of the temperature measurement.
  3. [Abstract and Section 5] The abstract calls the system 'self-contained,' but Section 3.1 explicitly states that 'A rotary vacuum pump connected to its base maintained the rough vacuum,' and Section 5 concedes that 'The need for a vacuum pump would be eliminated by use of a sealed, gas filled system.' As written, the demonstrated system requires an external vacuum pump and AC mains fans, so the portability/self-contained claim is not yet demonstrated. Please either temper the abstract and conclusion to say the cooling is self-contained while the full detector system remains a proof of principle, or add measurements with a sealed, gas-filled enclosure if one was tested.
minor comments (4)
  1. [Abstract] There is a typographical error: 'spectrsocopy' should be 'spectroscopy.'
  2. [Section 4.2 and 4.3, Figures 3 and 4] The spectra are shown as 'Normalized Count' but the text says only that channels were normalized around the 662 keV peak and counts were divided by the peak count. Please state the bin width and whether any smoothing was applied, since the line shapes appear smooth on the log scale.
  3. [Section 3.3, Table 1] The 'Electric 0.22' heat load for Section I is not explained in the text; clarify whether this is the Joule heating of the SiPM bias/readout electronics and how it was estimated.
  4. [References] Reference [5] (Moutinho et al.) is cited as a scintillating optical fiber dosimeter paper, but the URL provided points to a different article; please verify and correct the link.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper is a direct experimental measurement with an external PMT benchmark; the qualitative PMT comparison is under-supported but not circular.

full rationale

The paper is a direct experimental report, not a derivation. The dark-count reduction in Table 2 is an observed quantity, and the cooled SiPM spectra are measurements using a standard linear energy calibration from known photopeak energies; no fitted parameter is renamed as a prediction. The core comparative claim is benchmarked against a PMT run in the same setup, so the SiPM performance is externally anchored rather than derived from the paper's own assumptions. Statements such as the cooled SiPM being 'nearly identical' to a PMT and the PMT having superior noise below about 25 keV are qualitative and lack quantitative PMT metrics, but that is a missing-support or correctness concern, not circularity. There are no load-bearing self-citations; the cited references are datasheets and external SiPM characterizations that do not incorporate the present result. Therefore no circular step exists.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on measured temperatures, rates, and spectra plus background thermal and component data. No free parameters are fitted to achieve the reported result. The most consequential assumptions are datasheet accuracy, the quoted dark-count temperature scaling, and the adequacy of the linear energy calibration.

assumptions (5)
  • domain assumption Dark count of SiPMs falls by about one order of magnitude per 30 K temperature decrease.
    Section 2.2 cites Acerbi et al.; the cooling target is selected on this basis.
  • domain assumption Manufacturer datasheet values for the SiPM, TEC, and heat sink are accurate enough for design.
    Sections 2.2, 3.2, and 3.4 use SensL dark current and gain, Marlow XLT2422 maximum delta-T, and Wakefield-Vette thermal resistance as inputs.
  • standard math Thermal conductivity of gases is nearly independent of pressure down to below 1 mTorr, so 500 mTorr rough vacuum provides adequate insulation.
    Section 3.1 uses this physical approximation to justify rough vacuum instead of high vacuum.
  • domain assumption A linear energy calibration based on Cs-137 and Ba-133 photopeaks is valid across the measured energy range.
    Section 4.2 assigns keV values; any nonlinearity would shift the low-energy limit stated as approximately 25 keV.
  • domain assumption The room-temperature PMT measurement is a fair benchmark because the same NaI crystal and readout chain are used.
    Section 4.2 alternates the PMT in the same setup; differences in optical coupling geometry could bias the comparison.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Self-Contained, Cooled SiPM Array for Scintillation Spectroscopy." pith.science (2026). https://pith.science/paper/NSZGHS3J

@misc{pith2026190808852,
  author       = {Pith},
  title        = {Pith review of: Self-Contained, Cooled SiPM Array for Scintillation Spectroscopy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NSZGHS3J}},
  note         = {Machine review of arXiv:1908.08852}
}
read the original abstract

A simple, self-contained, thermo-electrically cooled SiPM system is presented which cools a SiPM array to -20 C. The array views a NaI scintillator through a 75 mm diameter glass window. Waste heat is removed with a large heat sink and AC fans. Above 40 keV in an air-coupled 2" x 2" NaI scintillator, the SiPM dark count rate was reduced by a factor of about 1000 when cooled. Spectroscopic performance when cooled was very similar to a PMT tested in the same setup, and adequate for nuclear spectrsocopy above 25 keV. Originally, water cooling was used but it was replaced by air cooling which is more suitable for a self-contained system, giving the advantage of portability without degrading performance. Straightforward improvements would allow cooling the SiPM to -30 C or below, which would further reduce the dark count rate and extend the spectroscopically useful range to even lower energies. Such a system would be rugged and suitable for field use, for instance for inspection of cargo for gamma ray emissions.

Figures

Figures reproduced from arXiv: 1908.08852 by the authors.

Figure 1
Figure 1. Side view of the cooling system designed for the SiPM. The thermal sections referred to in [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. On the left, the present device shown with the borosilicate viewing window at top right and the SiPM array visible below it. Limited space on the lower vacuum flange required the copper heat path to be offset as shown. While this interfered with efficient removal of heat by the external TECs, the offset would be absent in a more advanced design. One of the fans is visible as the black object at left front. On the ri… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

10 extracted references · 9 canonical work pages

  1. [1]

    C Series Datasheet

    “C Series Datasheet.” Sensl, Apr. 2018. https://www.onsemi.com/PowerSolutions/product.do?id=C- SERIES%20SIPM

    work page 2018

  2. [2]

    ArrayB-300Series-144P Preliminary Datasheet

    “ArrayB-300Series-144P Preliminary Datasheet.” Sensl, May 2013. Downloaded Feb. 2013; no longer available, superseded by C-Series

    work page 2013

  3. [3]

    AB424T-ARRAY144P

    “AB424T-ARRAY144P.” AiT Instruments. N.d. www.aid- instruments.com/AB424T_ARRAY144P_p/ab424tarray144p.htm

    work page

  4. [4]

    B-Series Preliminary Datasheet

    “B-Series Preliminary Datasheet.” Sensl, Feb. 2013. Downloaded Feb. 2013; no longer available, superseded by C-Series

    work page 2013

  5. [5]

    Development of a Scintillating Optical Fiber Dosimeter with Silicon Photomultipliers

    Moutinho, L.M., et al. “Development of a Scintillating Optical Fiber Dosimeter with Silicon Photomultipliers.” Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 2011, doi:10.1016/j.nima.2011.11.069. https://www.sciencedirect.com/science/article/pii/S0168900213013788 - 9 -

    work page doi:10.1016/j.nima.2011.11.069 2011

  6. [6]

    Cryogenic Characterization of FBK HD Near-UV Sensitive SiPMs

    Acerbi, Fabio, et al. “Cryogenic Characterization of FBK HD Near-UV Sensitive SiPMs.” IEEE Transactions on Electron Devices, vol. 64, no. 2, 2017, pp. 521–526., doi:10.1109/ted.2016.2641586. https://arxiv.org/pdf/1610.01915.pdf

    arXiv 2017

  7. [7]

    Technical Data Sheet for XLT2422

    “Technical Data Sheet for XLT2422.” Marlow Industries. N.d. https://www.digikey.com/product- detail/en/marlow-industries-inc/XLT2422-01LS/1681-1081-ND/6159129

    work page arXiv

  8. [8]

    Chhabra, R. P. CRC Handbook of Thermal Engineering. 2nd ed., Taylor & Francis, CRC Press, 2017

    work page 2017

Show all 10 references
  1. [9]

    Wakefield-Vette 392 Series

    “Wakefield-Vette 392 Series.” Wakefield-Vette. N.d. https://www.digikey.com/product- detail/en/wakefield-vette/392-300AB/345-1174-ND/4864908

  2. [10]

    Orion Fans OA109 ‘XC Series.’

    “Orion Fans OA109 ‘XC Series.’” Knight Electronics, Inc. N.d. https://orionfans.com/series.php?k=21

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.