Pith. sign in

REVIEW 4 major objections 4 minor 23 references

A Study of Afterglow Signatures in NaI and CsI Scintillator Modules for the Background and Transient Observer Instrument on COSI

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper shows CsI(Tl) afterglow lasts 5.6–8.6 times as long as NaI(Tl) after heavy-ion hits, and concludes NaI is the better detector for BTO because it loses only 0.02 s per orbit to afterglow versus ~9 s for CsI.

desk verdict Careful new afterglow measurements for SiPM-read Scionix modules, but the orbit-loss numbers rest on a threshold treatment that mixes incident and deposited energy. read the letter →

arxiv 2501.16434 v1 pith:SE6G6K6N submitted 2025-01-27 astro-ph.IM physics.ins-det

Hannah Gulick , Hiroki Yoneda , Tadayuki Takahashi , Claire Chen , Kazuhiro Nakazawa , Shunsaku Nagasawa , Mii Ando , Keigo Okuma
show 10 more authors
Alyson Joens Samer Al Nussirat Yasuyuki Shimizu Kaito Fujisawa Takayoshi Kohmura Kouichi Hagino Hisashi Kitamura Andreas Zoglauer Juan Carlos Martinez Oliveros John A. Tomsick
This is my paper · ORCID
classification astro-ph.IMphysics.ins-det
keywords gamma-rayscintillatorsscintillatorafterglowall-skysurveytime-domainastrophysicsNaI(Tl)Cssiliconphotomultipliersdetectordeadtime
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

The paper aims to settle a detector choice for the Background and Transient Observer (BTO), a gamma-ray transient monitor flying on the COSI mission. By irradiating NaI(Tl) and CsI(Tl) scintillators with helium and carbon beams, it measures the afterglow pulses that follow large energy deposits. It finds CsI's afterglow lasts 8.6 times (carbon beam) and 5.6 times (helium beam) as long as NaI's, and that CsI's stronger afterglow would force either a much longer dead time per event or a much higher trigger threshold. Combining the measured durations with simulated orbital background rates, the paper estimates that CsI would lose about 9 seconds of observing time per orbit to afterglow, while NaI would lose about 0.02 seconds. On this basis it concludes that NaI is the better scintillator for BTO.

What carries the argument

The central object is the afterglow signature itself: the train of small, delayed scintillation pulses that follows a saturating energy deposit. The paper's key measurement procedure is waveform analysis in which the post-pulse trend (scintillation decay plus electronics baseline) is fitted with a series of exponential functions, subtracted, and the residual voltage spread (sigma) is tracked in time bins until it returns to the pre-pulse baseline. The afterglow duration is defined as the time for that spread to reach baseline; this duration is then used as the dead time per afterglow-inducing event. A second load-bearing piece is the energy threshold for afterglow induction, estimated as 920 MeV for NaI (the helium-beam energy, since He produced essentially no NaI afterglow) and 164 MeV for CsI via a polynomial extrapolation constrained by a 662 keV no-afterglow calibration. These thresholds convert simulated hadronic background counts into afterglow event rates.

What would settle it

Measure afterglow in NaI(Tl) for deposited energies between roughly 100 and 900 MeV (for example by placing absorbers in front of a heavy-ion beam or using a tuneable beam energy). If NaI shows afterglow durations well above zero at energies below 920 MeV, the assumed 920 MeV threshold is wrong and the NaI dead-time loss of 0.02 s per orbit is an underestimate; the trade-study conclusion would need revisiting. Alternatively, a flight measurement of afterglow trigger rates inside the South Atlantic Anomaly could test the rate prediction directly.

Watch

Extended reading notes

Core claim

On its own terms, this paper establishes that both NaI(Tl) and CsI(Tl) detector modules produce measurable afterglow after heavy-ion irradiation, but with dramatically different severities. For a 350 MeV/u carbon beam, afterglow pulses persist for a median 2.40 ms in CsI versus 0.28 ms in NaI; for a 230 MeV/u helium beam, 0.9 ms versus 0.16 ms. The CsI afterglow is also more intense, with a residual voltage spread about twice that of NaI even though NaI has a higher gain. Because afterglow events look like small pulses that can falsely trigger a gamma-ray detector, the paper translates these durations into dead-time costs: after estimating that a BTO-like instrument in low Earth orbit would see afterglow-inducing particles every ~74 s in NaI and every ~1.4 s in CsI, it finds the observing-time loss per orbit is ~0.02 s for NaI and ~9 s for CsI. The paper's conclusion is that NaI is the better choice for BTO despite CsI's higher light yield and radiation hardness.

Load-bearing premise

The comparison rests on assuming that only particles above a particular energy threshold (920 MeV in NaI, 164 MeV in CsI) produce afterglow and that each such event costs exactly the measured carbon-beam afterglow duration in dead time; if significant afterglow occurs at lower energies, or if the South Atlantic Anomaly contributes more events than the simulation assumes, the 0.02 s versus 9 s loss comparison and the choice of NaI could change.

Editorial extensions

If this is right

  • If NaI is used on BTO, afterglow-inducing events are expected every ~74 seconds per detector, meaning afterglow dead time costs about 0.02 s per orbit.
  • If CsI were used instead, afterglow-inducing events would arrive every ~1.4 seconds, and 2.4 ms dead time per event would cost about 9 s per orbit, roughly 400 times more lost time.
  • CsI afterglow would force either a trigger threshold about 3 times higher or a dead time about 7 times longer than NaI to avoid false triggers, both of which degrade the 30 keV–2 MeV science band.
  • These results imply that any space-based gamma-ray instrument using CsI(Tl) with SiPM readout should plan for afterglow mitigation such as post-saturation dead time or coincidence requirements.

Reading between the lines

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

  • The measured energy thresholds and durations suggest a testable scaling law: afterglow duration in these crystals likely grows with total deposited energy, and the polynomial fit used for CsI could be extended to NaI to predict its afterglow at energies between a few hundred MeV and 1 GeV, a regime more typical of trapped protons than the heavy-ion beams used here.
  • Because trapped protons in the South Atlantic Anomaly have lower energies than the carbon beam but much higher flux, the orbital loss estimate may be dominated by how often a 100–500 MeV proton deposit crosses the threshold; the paper's threshold method could be sharpened with proton-beam measurements.
  • A similar trade-study logic applies to other scintillator-based transient monitors: the choice between high light yield and long afterglow can be framed as an observing-time budget, where the product of afterglow duration and event rate determines which material wins.
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.

Referee Report

4 major / 4 minor

Summary. The paper reports measurements of afterglow in NaI(Tl) and CsI(Tl) scintillator modules with SiPM readout, irradiated with 230 MeV/u He and 350 MeV/u C beams at HIMAC. The authors measure afterglow durations of 2.40 ± 0.5 ms (CsI, C) and 0.28 ± 0.07 ms (NaI, C), with corresponding He values of 0.9 ± 0.2 ms and 0.16 ± 0.04 ms, and use these to argue that CsI afterglow is 8.6× and 5.6× longer than NaI for C and He beams, respectively. They then use MEGAlib simulations of a BTO-like low-Earth orbit to estimate afterglow-inducing event rates, assign the measured C-beam durations as dead times, and conclude that NaI loses ~0.02 s per orbit to afterglow versus ~9 s for CsI, motivating the choice of NaI for BTO. The direct duration measurements are repeated over 50+ pulses per configuration with gain correction and error bars, but the orbital impact calculations rely on several assumptions about afterglow thresholds and dead times that are not robustly justified.

Significance. If the measured afterglow contrast is correct, the paper provides a useful, quantitatively documented input for scintillator selection in space missions using SiPM readout, and the HIMAC dataset with per-pulse statistics is a strength. The central qualitative conclusion—that CsI exhibits substantially stronger and longer afterglow than NaI under heavy-ion irradiation—is well supported by the waveform analysis. However, the quantitative claims that drive the trade-study conclusion (event spacings of ~70 s vs ~1.4 s and lost time of 0.02 s vs 9 s per orbit) depend on threshold and dead-time choices that are not securely grounded, so the paper's main quantitative predictions need revision before the conclusion can be considered robust.

major comments (4)
  1. [§3.1] The afterglow-inducing thresholds are defined as incident kinetic energies (920 MeV for NaI, 164 MeV for CsI) and used to integrate the MEGAlib hadronic flux, but the HIMAC data establish afterglow as a function of energy deposited in the crystal: the 920 MeV He beam deposits only 275 MeV in NaI and 325 MeV in CsI, while the C beam deposits 1.65/2 GeV. A 920 MeV proton traversing a 3.8-cm crystal deposits far less energy than the 920 MeV He used in the experiment, so integrating all hadrons above 920 MeV likely misstates the event rate. The authors should apply the energy cut on deposited energy in the scintillator (as they do for the 74.5 MeV CsI validation) or explicitly justify why incident kinetic energy is the relevant variable; as written, the claimed event spacings and the 0.02 s vs 9 s lost-time comparison are not securely grounded.
  2. [§3.1] The CsI threshold of 164 MeV is obtained by extrapolating the afterglow-duration-versus-excitation-energy relation from the same HIMAC data (including the 662 keV calibration point) and is then used to compute the CsI afterglow event rate of 0.7 particles/s. This makes the 'every 1.4 s' rate partly an encoding of the measured durations rather than an independent prediction, and the uncertainty in the threshold from the polynomial fit is not propagated. The paper should present this as an explicit calibration step and propagate the threshold uncertainty into the event-rate and lost-time estimates, or use an independently justified threshold.
  3. [§3.1 and §3.2] The dead-time-per-event assignment uses the C-beam afterglow durations (2.4 ms for CsI, 0.28 ms for NaI) for all events above threshold, even though the HIMAC data show a strong dependence on deposited energy (the He beam gives 0.9 ms in CsI). If many events just above threshold deposit less energy, the assigned dead time overestimates the lost time; if the threshold is set too low, the event rate may be overestimated. A sensitivity study over reasonable threshold and dead-time choices is needed to support the 0.02 s vs 9 s conclusion, which is the central quantitative basis for the trade-study decision.
  4. [§3.1] The rate calculation explicitly excludes the South Atlantic Anomaly ('outside of the SAA'), yet the text states that afterglow in the SAA 'will be strong' and that BTO will remain powered during SAA crossings. Since trapped hadrons in the SAA may dominate the afterglow-inducing event rate, the quoted lost-time estimates are incomplete as stated. The authors should either quantify the SAA contribution or explicitly restrict the conclusion to non-SAA portions of the orbit.
minor comments (4)
  1. [§3.2] There is an internal inconsistency in the CsI event spacing: §3.1 reports 0.7 particles/s (one event every 1.4 s), while §3.2 states 'every ~0.7 seconds in the CsI detector'; the latter should read ~1.4 s.
  2. [§3.1] The NaI threshold is justified by saying the He afterglow signal 'can be approximated as zero,' but Table 2 lists a nonzero NaI He afterglow duration of 0.16 ± 0.04 ms; this approximation should be stated more carefully, and its effect on the rate estimate should be addressed.
  3. [§3.1] The claim that the lost observing time in NaI is '~0.01×' that in CsI is numerically inconsistent with the quoted values 0.02 s vs 9 s, which give a ratio of ~0.002.
  4. [Figure 4] The caption refers to a 'post-post residual' in the third panel; this appears to be a typo for 'post-pulse residual.'

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the afterglow durations are direct measurements and the orbital lost-time estimate combines them with an independent MEGAlib/SPENVIS flux simulation.

full rationale

The derivation chain is not circular. The central measured quantities—afterglow durations of 0.28 ms (NaI, C), 2.4 ms (CsI, C), 0.16 ms (NaI, He), 0.9 ms (CsI, He)—are direct waveform data from the HIMAC campaign (Section 2.3, Table 2), not outputs of any model. The CsI afterglow threshold of 164 MeV is calibrated from those same measurements by fitting excitation energy versus afterglow duration with a 662 keV no-afterglow constraint, and it is cross-checked externally against Rau et al. (2005) through a MEGAlib deposited-energy calculation. Using that calibrated threshold to integrate an independent SPENVIS/MEGAlib orbital hadronic flux, then multiplying by the measured C-beam dead times to estimate lost observing time, is a forward application of calibrated parameters rather than a restatement of the fit: the 0.02 s versus 9 s comparison depends on the simulated flux ratio above the two thresholds, which is not fixed by the afterglow measurements alone. The paper does contain an internal inconsistency in the quoted CsI event spacing (Section 3.1 says every 1.4 s; Section 3.2 says every ~0.7 s) and the choice of incident-energy thresholds is debatable, but these are correctness and modeling concerns, not circularity. No load-bearing self-citation appears; citations to the authors' prior work (Gulick et al. 2024, Tomsick et al. 2023) provide only detector geometry and mission context. The core conclusion is therefore independently grounded in the HIMAC data plus an external orbital background simulation.

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

The afterglow durations are measured, but the orbit-level rate comparison rests on several hand-set thresholds and an extrapolated energy cutoff. No new entities are introduced.

free parameters (4)
  • Afterglow duration detection thresholds = 15-bin mean, 1-sigma, 3-sigma, 0.07 ms buffer
    Hand-tuned criteria in Section 2.3 define when afterglow is declared undetectable; changing these changes measured durations.
  • NaI afterglow-inducing energy threshold = 920 MeV
    Set equal to total He excitation energy in Section 3.1 because NaI He afterglow was treated as approximately zero; used to compute NaI event rate.
  • CsI afterglow-inducing energy threshold = 164 MeV
    Extrapolated from a second-order polynomial fit to the two HIMAC beam energies plus a 662 keV no-afterglow constraint in Section 3.1; used to compute CsI event rate.
  • Afterglow dead time per event = 0.28 ms (NaI), 2.4 ms (CsI)
    Assigned as the C-beam afterglow durations from Table 2 in Section 3.1; directly sets the per-orbit lost time estimates.
assumptions (4)
  • domain assumption Residual voltage spread in the post-pulse waveform is a valid proxy for afterglow pulses once the general trend is removed.
    Section 2.2 assumes afterglow appears as broadening of the voltage distribution; noise and electronics effects are treated as removable by the exponential fits.
  • domain assumption Afterglow-inducing energy thresholds measured or estimated for cylindrical HIMAC modules also apply to the rectangular flight-model BTO crystals.
    Section 3.1 states the afterglow-inducing energy is intrinsic to the scintillator material, but the flight detectors differ in geometry and housing.
  • domain assumption SPENVIS AP9 90th percentile trapped-particle fluxes and the Ajello and Mizuno background models adequately represent the BTO orbit outside the SAA.
    Section 3.1 uses these models without propagated uncertainties.
  • ad hoc to paper An event's total excitation energy determines whether it induces afterglow, with a sharp threshold and no dependence on particle type or angle.
    Section 3.1 applies a single energy cutoff to all hadronic and trapped-hadronic background components.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Study of Afterglow Signatures in NaI and CsI Scintillator Modules for the Background and Transient Observer Instrument on COSI." pith.science (2026). https://pith.science/paper/SE6G6K6N

@misc{pith2026250116434,
  author       = {Pith},
  title        = {Pith review of: A Study of Afterglow Signatures in NaI and CsI Scintillator Modules for the Background and Transient Observer Instrument on COSI},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SE6G6K6N}},
  note         = {Machine review of arXiv:2501.16434}
}
read the original abstract

We present measurements of the afterglow signatures in NaI(Tl) and CsI(Tl) detector modules as part of the Background and Transient Observer (BTO) mission detector trade-study. BTO is a NASA Student Collaboration Project flying on the Compton Spectrometer and Imager (COSI) Small Explorer mission in 2027. The detectors utilized in this study are cylindrical in shape with a height and diameter of 5.1 cm and were read out by silicon photomultipliers (SiPMs). We conducted a radiation campaign at the HIMAC accelerator in Japan where the scintillators were irradiated with a 230 MeV/u helium beam (He beam) and 350 MeV/u carbon beam (C beam). We find that both the CsI and NaI scintillators exhibit afterglow signatures when irradiated with the C and He beams. The CsI crystal exhibits a stronger afterglow intensity with afterglow pulses occurring for an average 2.40 ms for C and 0.9 ms for He after the initial particle pulse. The duration of afterglow pulses in CsI is 8.6x and 5.6x the afterglow signal duration in NaI for C and He (0.28 ms and 0.16 ms, respectively). Although CsI has advantages such as a higher light yield and radiation hardness, the stronger afterglows in the CsI detector increase the complexity of the electronics and lead to a ~7x larger dead time per afterglow event or a ~3x higher energy threshold value. We use the measured dead times to predict the amount of observing time lost to afterglow-inducing events for an instrument like BTO in low Earth orbit. We simulate the background rates in a BTO-like orbit and find a total value of 114 counts/s for the full two-detector system. Based on the particle energies in the HIMAC experiment, we then determine that an event with sufficient energy to produce an afterglow signal occurs once every ~70 s and ~1.4 s in NaI and CsI detectors, respectively. Thus, we conclude that NaI is the better choice for the BTO mission.

Figures

Figures reproduced from arXiv: 2501.16434 by the authors.

Figure 1
Figure 1. The Scionix cylindrical detector schematic. The detector contains [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The HIMAC experimental setup. The Scionix detector (NaI or CsI) [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. An example waveform taken with the CsI detector and the C beam. The time-axis values are chosen to highlight the pulse’s features including the [left] [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: (Top) An example of a waveform taken with the CsI detector with SiPM readout when irradiated with the C beam. The region classification used in this [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The post-pulse residuals (black) for a single pulse in all four experimental setups. The residuals are found by subtracting a fit to the general trend of the [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: The estimated background rates for the BTO system including two [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 22 canonical work pages

  1. [1]

    2008, The Astrophysical Journal, 689, 666

    Ajello, M., Greiner, J., Sato, G., et al. 2008, The Astrophysical Journal, 689, 666

    work page 2008

  2. [2]

    2009, Nuclear In- struments and Methods in Physics Research Section A: Accelerators, Spec- trometers, Detectors and Associated Equipment, 606, 585

    Alfassi, Z., Ifergan, Y ., Wengrowicz, U., & Weinstein, M. 2009, Nuclear In- struments and Methods in Physics Research Section A: Accelerators, Spec- trometers, Detectors and Associated Equipment, 606, 585

    work page 2009

  3. [3]

    W., Payne, S

    Bizarri, G., Moses, W. W., Payne, S. A., & Williams, R. T. 2011, in Hard X-Ray, Gamma-Ray, and Neutron Detector Physics XIII, ed. L. A. Franks, R. B

    work page 2011

  4. [4]

    S., Fishman, G

    Briggs, M. S., Fishman, G. J., Connaughton, V ., et al. 2010, Journal of Geo- physical Research (Space Physics), 115, A07323 10

    work page 2010

  5. [5]

    2022, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 513, 33

    Dilillo, G., Zampa, N., Campana, R., et al. 2022, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 513, 33

    work page 2022

  6. [6]

    R., Smith, D

    Dwyer, J. R., Smith, D. M., & Cummer, S. A. 2012, Space Science Reviews, 173, 133

    work page 2012

  7. [7]

    R., 1982, IEEE Transactions on Nuclear Science, 29, 1237

    Farukhi, M. R., 1982, IEEE Transactions on Nuclear Science, 29, 1237

    work page 1982

  8. [8]

    J., Bhat, P

    Fishman, G. J., Bhat, P. N., Mallozzi, R., et al. 1994, Science, 264, 1313

    work page 1994

Show all 23 references
  1. [9]

    W., Smith, D

    Grefenstette, B. W., Smith, D. M., Hazelton, B. J., & Lopez, L. I. 2009, Journal of Geophysical Research (Space Physics), 114, A02314

    2009

  2. [10]

    S., et al

    Guiriec, S., Connaughton, V ., Briggs, M. S., et al. 2011, The Astrophysical Journal Letters, 727, L33

    2011

  3. [11]

    C., Neights, E., Al Nussirat, S., et al

    Gulick, H. C., Neights, E., Al Nussirat, S., et al. 2024, arXiv e-prints, arXiv:2407.07155

    2024 arXiv

  4. [12]

    2022, Crystals, 12

    Hawrami, R., Ariesanti, E., Farsoni, A., Szydel, D., & Sabet, H. 2022, Crystals, 12

    2022

  5. [13]

    M., & Beloborodov, A

    Kaspi, V . M., & Beloborodov, A. M. 2017, Annual Review of Astronomy and Astrophysics, 55, 261

    2017

  6. [14]

    Koppert, W. J. C., van der Velden, S., Steenbergen, J. H. L., & de Jong, H. W. A. M. 2018, Physics in Medicine and Biology, 63, 065006

    2018

  7. [15]

    Lecoq, P., 2020, Scintillation Detectors for Charged Particles and Photons, (Cham: Springer International Publishing), 45–89

    2020

  8. [16]

    S., Paciesas, W

    Mallozzi, R. S., Paciesas, W. S., Pendleton, G. N., et al. 1995, The Astrophysi- cal Journal, 454, 597

    1995

  9. [17]

    2004, The Astrophysical Journal, 614, 1113 Pe’er, A., M ´esz´aros, P., & Rees, M

    Mizuno, T., Kamae, T., Godfrey, G., et al. 2004, The Astrophysical Journal, 614, 1113 Pe’er, A., M ´esz´aros, P., & Rees, M. J. 2006, The Astrophysical Journal, 642, 995

    2004

  10. [18]

    V ., Hurley, K., & Lichti, G

    Rau, A., Kienlin, A. V ., Hurley, K., & Lichti, G. G. 2005, The Astronomy & Astrophysics Journal, 438, 1175

    2005

  11. [19]

    B., et al

    Ryde, F., Axelsson, M., Zhang, B. B., et al. 2010, The Astrophysical Journal Letters, 709, L172

    2010

  12. [20]

    M., Lopez, L

    Smith, D. M., Lopez, L. I., Lin, R. P., & Barrington-Leigh, C. P. 2005, Science, 307, 1085

    2005

  13. [21]

    A., Boggs, S

    Tomsick, J. A., Boggs, S. E., Zoglauer, A., et al. 2023, arXiv e-prints, arXiv:2308.12362

    2023 arXiv

  14. [22]

    D., Rooney, B

    Valentine, J. D., Rooney, B. D., & Dorenbos, P. 1998, IEEE Transactions on Nuclear Science, 45, 1750

    1998

  15. [23]

    2006, New Astronomy Reviews, 50, 629 11

    Zoglauer, A., Andritschke, R., & Schopper, F. 2006, New Astronomy Reviews, 50, 629 11

    2006

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.