Measured gain suppression in FBK LGADs with different active thicknesses

A. Bisht; A. Molnar; A. Seiden; B. Schumm; F. McKinney-Martinez; G. Paternoster; J. Ding; J. Ott; J. Yang; M. Boscardin

REVIEW 2 major objections 4 minor 12 references

Measured gain suppression in FBK LGADs with different active thicknesses

T0 review · 2 major / 4 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Gain suppression in LGADs peaks at 40–60 degrees for MeV proton deposits, growing with bias voltage.

desk verdict A modest, honest LGAD gain-suppression measurement for FBK sensors with a real normalization caveat on the 100 µm device, but the qualitative angle trend holds up. read the letter →

arxiv 2502.02244 v2 pith:L4Q4M3ZH submitted 2025-02-04 physics.ins-det hep-ex

J. Yang , S. Braun , Q. Buat , J. Ding , M. Harrison , P. Kammel , S.M. Mazza , F. McKinney-Martinez

show 11 more authors

A. Molnar J. Ott A. Seiden B. Schumm Y. Zhao Y. Zhang V. Tishchenko A. Bisht M. Centis-Vignali G. Paternoster M. Boscardin

This is my paper · ORCID

classification physics.ins-dethep-ex

keywords low-gainavalanchedetectorLGADgainsuppressionchargedensityprotonbeamenergylinearityactivetargetparticleidentification

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 reports measurements of how the internal gain of low-gain avalanche detectors (LGADs) responds to dense deposits of charge from a tuned proton beam, using sensors with 50, 100, and 150 micrometers of active thickness. The central result is that for MeV-range protons, the measured gain is far below the gain the same detectors show for minimum-ionizing particles, and the shortfall grows as the bias voltage is increased. The suppression also changes with the angle of incidence: it gets stronger up to roughly 40–60 degrees and then weakens, because at large angles the proton's Bragg peak deposits more charge close to the gain layer. The authors conclude that gain suppression is a serious nonlinearity for any experiment using LGADs as an active target for energy measurement, because particle identification relies on a linear energy response.

What carries the argument

The central object is the highly doped gain layer of an LGAD, the thin region where avalanche multiplication begins. The mechanism being studied is gain suppression: a dense cloud of ionization from a stopping proton drifts into the gain layer and, once multiplied, partially shields the electric field that sustains the avalanche, so the effective gain falls below its small-deposit value. The measurements trace this suppression by dividing the charge collected in an LGAD by the charge collected in a geometrically identical PIN diode without a gain layer, varying the bias voltage, proton energy, and angle of incidence; the shallow gain layer of the tested devices is the property they single out to explain the strength of the observed effect.

What would settle it

Measure the active thickness of each reference PIN with an independent technique—capacitance–voltage profiling or alpha-particle energy loss—and recompute the gain from the raw LGAD and PIN charge data. If the 100 micrometer PIN is truly 80 micrometers, the absolute gain values for that sensor shift; the central claim survives only if the corrected gains still show the same strong bias-dependent suppression and the same angular peak at intermediate angles across all three thicknesses.

Watch

Extended reading notes

Core claim

The paper's central claim is that all tested LGADs compress their gain substantially when a MeV proton deposits a high density of charge, and that this compression is governed by bias voltage, incidence angle, and proton energy. The gain at a given bias is not constant: it drops sharply from its minimum-ionizing value, and the drop is biggest at intermediate angles of incidence rather than at perpendicular incidence. The authors attribute the angular behavior to two competing effects: angled tracks spread the charge across the gain layer, which reduces the local charge density and lessens suppression, while at large angles more of the Bragg-peak charge is deposited near the gain layer, which increases suppression because charge generated close to the gain layer is less effective at producing gain. They further note that the tested devices, which have a very shallow gain layer, show stronger suppression than comparable sensors with a different doping profile, suggesting that the compact high-field region amplifies the shielding of the multiplication field by the moving charge cloud.

Load-bearing premise

The gain is defined as the charge collected by an LGAD divided by the charge collected by a same-thickness PIN detector, which assumes the PIN has the same active thickness and charge-collection efficiency as the LGAD apart from the gain layer; the paper reports that its 100 micrometer PIN behaves as if it were 80 micrometers thick, so the normalization for that sensor is uncertain and the reported gains carry an unquantified systematic shift.

Editorial extensions

If this is right

For any dense deposit, the LGAD gain is a strong nonlinear function of deposited charge, so an active target must apply per-pulse, angle-dependent gain corrections to recover linear energy response.
Raising the bias voltage to increase gain for minimum-ionizing tracks amplifies the suppression for MeV protons, forcing a trade-off between timing resolution and energy linearity.
The suppression pattern means a stopping particle and a through-going particle of the same kinetic energy will register different charge, so particle identification based on energy deposits requires modeling the Bragg-peak position.
Because the angular peak of suppression lies near 40–60 degrees, tracks at these angles in a detector will be the most distorted if the calibration is done only at normal incidence.
The comparison between gain-layer profiles indicates that the doping depth of the gain layer is a design handle for reducing suppression, since a shallower gain layer is associated with stronger nonlinearity.

Reading between the lines

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

A quantitative model could be built from the data: plot the measured suppression against the charge density projected on the gain layer, accounting for the proton's stopping profile and lateral diffusion; if a single critical density triggers the suppression, the angular peak at 40–60 degrees should emerge naturally from the geometry.
The same mechanism should affect the timing resolution of LGADs for highly ionizing particles, because the leading-edge signal depends on the amount and arrival time of charge at the gain layer; the paper does not measure timing, so this is a testable extension.
If a deeper gain layer reduces suppression, then devices currently being developed for large timing detectors could be tuned toward better linearity at the cost of some time resolution; this trade-off is a design question the paper's comparison motivates but does not resolve.
The 100 micrometer normalization problem implies that any absolute gain comparison between this dataset and other measurements should be treated with caution until the PIN thickness is independently verified.

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

Quick take: this is a clean, useful follow-up measurement of LGAD gain suppression for FBK sensors. The genuinely new content is the systematic angle, energy, and thickness dependence for 50, 100, and 150 µm devices, plus the direct comparison to their earlier HPK data. The measurement approach is straightforward — RBS proton beam, PIN normalization, 90Sr MIP reference — and the paper is honest about its main limitation: the 100 µm PIN behaves as if it were 80 µm thick.

What it does well: the experimental campaign is careful, the data cover a useful parameter space for the PIONEER active target, and the reported qualitative trend — suppression grows with bias and with angle up to 40–60°, then drops at large angles due to charge deposited near the gain layer — is clearly supported by the figures. The discussion of the shallow FBK gain layer as a likely cause of stronger suppression than HPK is reasonable and appropriately labeled as a hypothesis.

Soft spots, in proportion: the lack of error bars on any of the gain values is a real gap. Some runs are excluded for amplifier saturation, and the reader doesn't get a sense of run-to-run stability. The 100 µm PIN thickness discrepancy is acknowledged in Section 4 but not propagated into the quoted numbers. The stress-test analysis is right that this is not just a 20% scale error: for stopping protons, the missing last ~10 µm of silicon in the PIN removes part of the Bragg peak, so the denominator is too small in a way that depends on angle and energy. That means the 100 µm gain values, and any device-to-device comparison involving them, carry an unquantified systematic shift. The qualitative angle dependence for that device is still visible, and the 50 and 150 µm devices don't have this problem, so the central claim survives.

Verdict: the paper deserves peer review and likely publication after the authors add uncertainty estimates, clarify the 100 µm thickness calibration, and either renormalize or caveat that dataset. The reader's conditional verdict is about right; I'd be slightly more positive about the qualitative contribution.

Referee Report

2 major / 4 minor

Summary. The manuscript reports an experimental study of gain suppression in three FBK LGADs with active thicknesses of 50, 100, and 150 micrometers, using MeV-range protons from the CENPA tandem accelerator and a 90Sr MIP reference. Gain is defined as the ratio of collected charge in an LGAD to that in a matched PIN diode without a gain layer, measured as a function of bias voltage, incidence angle, and proton energy. The main finding is that all LGADs show substantial gain suppression relative to laboratory MIP gain, with a non-monotonic angular dependence: suppression increases up to 40-60 degrees and then decreases because of charge deposition near the gain layer. The authors also compare with earlier HPK measurements and attribute the stronger FBK suppression to a very shallow gain layer.

Significance. The paper provides a useful dataset for LGAD applications such as the PIONEER active target, where energy linearity for stopping particles is critical. The systematic variation of angle across three thicknesses is valuable, and the direct comparison with an external MIP benchmark is a strength. The paper does not rely on free parameters or fitting; the gain is directly measured. However, the quantitative gain values for the 100 micrometer device are unreliable without a corrected normalization, and the absence of error bars limits the precision of all reported numbers. The central qualitative claims are likely robust, but the paper needs revision to support its quantitative statements.

major comments (2)

[Sec. 4 and Fig. 1 (left); Sec. 5] The 100 micrometer PIN device (W8 PIN) is shown in Sec. 4 to collect charge consistent with an 80 micrometer active thickness, yet the gain in Sec. 5 for the 100 micrometer LGAD is computed as the ratio to this same PIN. For stopping protons, the PIN's missing ~20 micrometers of active silicon removes not only about 20% of the deposited charge but also the Bragg-peak tail near the end of the proton range; this missing fraction is angle- and energy-dependent because the geometric track length changes with angle. The resulting W8 gain values and any device-to-device comparisons in Sec. 6 therefore carry an unquantified systematic error. Please correct the normalization, propagate the systematic uncertainty, or explicitly restrict the quantitative claims for the 100 micrometer device.
[Fig. 2 and Sec. 5] No error bars or statistical uncertainties are shown for any of the gain points in Fig. 2, and the caption states that some runs were excluded because of amplifier or digitizer saturation. Since saturation occurs at the highest gains, the exclusion is correlated with the high-bias and large-angle points that are central to the claimed bias and angle trends. Please provide per-point uncertainties and a quantitative assessment of how the excluded runs affect the stated trends, or at least identify which points are affected and justify that the conclusions remain unchanged.

minor comments (4)

[Sec. 2] The first sentence says 'protons of 1.8 MeV, 3 MeV, etc. momenta were used'; since the listed quantities are kinetic energies, the wording should be 'energies' or 'kinetic energies' for consistency.
[Fig. 1 (left)] The y-axis of the collected-charge plot has no units stated in the text or caption; please add units (e.g., fC) to the figure or the caption.
[Sec. 4] The sentence 'The charge values reflect the PIN thickness at 50 um and 150 um (around 0.1 fC per 10 um)' would be clearer if the expected charge per micrometer were defined explicitly, since the reader must infer the conversion from energy loss.
[Sec. 6] The comparison with HPK sensors from Ref. [7] is made without stating whether those data were reanalyzed with the same analysis package and similar run selections; please add a sentence clarifying the consistency of the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the gain is a directly measured charge ratio against external PIN and MIP benchmarks, and the 100 µm PIN normalization issue is an acknowledged systematic uncertainty rather than a circular reduction.

full rationale

The paper's central quantity, 'gain', is defined operationally as the ratio of charge collected by an LGAD to charge collected by a same-nominal-thickness PIN diode without a gain layer, with the MIP response from a 90Sr source used as an external reference. No parameter is fitted to the proton data and then reused as a 'prediction'; the angle, bias, and energy trends in Section 5 are direct measurements of this ratio. The paper itself flags the one normalization concern: Section 4 states that the 100 µm PIN 'seems to behave as being 80µm thick,' so the W8 gain values carry an unquantified systematic shift. This is a calibration or systematic-error issue, not a circularity: the PIN response is measured, not derived from the LGAD gain, and the assumption of equal active thickness is a stated physical assumption that could fail without making the measurement tautological. Self-citations appear only for setup, previous measurement methodology, and prior empirical observations of X-ray depth effects; they are not used as a uniqueness theorem or to forbid alternative explanations. In particular, the comparison with HPK sensors in the Conclusions is an external empirical contrast, and the proposed explanation in terms of shallow gain-layer doping is explicitly framed as a hypothesis. No equation reduces to its own input, no fitted input is renamed as a prediction, and no known result is merely relabeled. The honest verdict is no significant circularity, with the 100 µm normalization issue recorded as a correctness risk rather than a circular step.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no free parameters or new entities. Its claims rest on standard detector physics assumptions: the gain-layer shielding mechanism, the PIN as unity-gain reference, and Bragg-peak energy deposition. The PIN-reference assumption is the most fragile and is partially contradicted by the paper's own 100 micrometer thickness observation.

assumptions (3)

domain assumption Gain suppression is triggered by shielding of the electric field in the gain layer caused by multiplication of charge carriers in the bulk.
Invoked in Section 1 as the mechanism explaining the observed dependence; taken from prior literature, not derived in this paper.
domain assumption The PIN diode response is a valid unity-gain reference for the LGAD of the same nominal thickness.
Central to the gain definition in Section 5; the paper itself questions this for the 100 micrometer device in Section 4.
domain assumption The proton energy deposition profile follows the Bragg peak as described by stopping power.
Used to interpret angle dependence and the proximity of charge deposition to the gain layer in Section 5.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Measured gain suppression in FBK LGADs with different active thicknesses." pith.science (2026). https://pith.science/paper/L4Q4M3ZH

@misc{pith2026250202244,
  author       = {Pith},
  title        = {Pith review of: Measured gain suppression in FBK LGADs with different active thicknesses},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L4Q4M3ZH}},
  note         = {Machine review of arXiv:2502.02244}
}

abstract

In recent years, the gain suppression mechanism has been studied for large localized charge deposits in Low-Gain Avalanche Detectors (LGADs). LGADs are a thin silicon detector with a highly doped gain layer that provides moderate internal signal amplification. Using the CENPA Tandem accelerator at the University of Washington, the response of LGADs with different thicknesses to MeV-range energy deposits from a proton beam were studied. Three LGAD prototypes of 50~$\mu$m, 100~$\mu$m, 150~$\mu$m were characterized. The devices' gain was determined as a function of bias voltage, incidence beam angle, and proton energy. This study was conducted in the scope of the PIONEER experiment, an experiment proposed at the Paul Scherrer Institute to perform high-precision measurements of rare pion decays. LGADs are considered for the active target (ATAR) and energy linearity is an important property for particle ID capabilities.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[1]

Pellegrini et al.,Technology developments and first measurements of Low Gain Avalanche Detectors (LGAD) for high energy physics applications,Nucl

G. Pellegrini et al.,Technology developments and first measurements of Low Gain Avalanche Detectors (LGAD) for high energy physics applications,Nucl. Instrum. Meth.A765(2014) 12 – 16

work page 2014
[2]

Comparison of 35 and 50 {\mu}m thin HPK UFSD after neutron irradiation up to 6*10^15 neq/cm^2

Y. Zhao et al.,Comparison of 35 and 50µm thin HPK UFSD after neutron irradiation up to 6*1015 neq/cm2,Nucl. Instrum. Meth. A924(2019) 387–393, [1803.02690]. [3]ATLAScollaboration, HGTD,Technical Design Report: A High-Granularity Timing Detector for the ATLAS Phase-II Upgrade, Tech. Rep. CERN-LHCC-2020-007. ATLAS-TDR-031, CERN, Geneva, Jun, 2020

work page Pith review arXiv 2019
[4]

CMS collaboration,A MIP Timing Detector for the CMS Phase-2 Upgrade, Tech. Rep. CERN-LHCC-2019-003. CMS-TDR-020, CERN, Geneva, Mar, 2019

work page 2019
[5]

An LGAD-based full active target for the PIONEER experiment

Electron-Ion collider, “https://www.bnl.gov/eic/.” [6]PIONEERcollaboration, S. M. Mazza,An LGAD-Based Full Active Target for the PIONEER Experiment,Instruments5(2021) 40, [2111.05375]

work page Pith review arXiv 2021
[7]

Gain suppression study on LGADs at the CENPA tandem accelerator

S. Braun et al.,Gain suppression study on LGADs at the CENPA tandem accelerator,Nucl. Instrum. Meth. A1064(2024) 169395, [2405.02550]

work page Pith review arXiv 2024
[8]

M. C. J.-R. et al.,Study of ionization charge density-induced gain suppression in LGADs,Sensors22 (2022)

work page 2022
[9]

Currás, M

E. Currás, M. Fernández and M. Moll,Gain reduction mechanism observed in Low Gain Avalanche Diodes,Nucl. Instrum. Meth. A1031(May, 2022) 166530

work page 2022
[10]

D. S. et al.,Status of and operating experience with the university of washington superconducting booster linac,Nucl. Instrum. Meth A 287, 247(1990)

work page 1990

Show all 12 references

[11]

R. P. et al.,Effect of deep gain layer and carbon infusion on LGAD radiation hardness,Journal of Instrumentation15(2020) P10003

2020
[12]

Breton, C

D. Breton, C. Cheikali, E. Delagnes, J. Maalmi, P. Rusquart and P. Vallerand,Fast electronics for particle Time-Of-Flight measurement, with focus on the SAMPIC ASIC,Nuovo Cim. C43(2020) 7

2020
[13]

S. M. Mazza et al.,Synchrotron light source X-ray detection with Low-Gain Avalanche Diodes, Journal of Instrumentation18(2023) P10006, [2306.15798]

2023 arXiv
[14]

Molnar et al.,Synchrotron light source focused X-ray detection with LGADs, AC-LGADs and TI-LGADs,2504.18638

A. Molnar et al.,Synchrotron light source focused X-ray detection with LGADs, AC-LGADs and TI-LGADs,2504.18638. – 6 –

arXiv

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.

[1] [1]

Pellegrini et al.,Technology developments and first measurements of Low Gain Avalanche Detectors (LGAD) for high energy physics applications,Nucl

G. Pellegrini et al.,Technology developments and first measurements of Low Gain Avalanche Detectors (LGAD) for high energy physics applications,Nucl. Instrum. Meth.A765(2014) 12 – 16

work page 2014

[2] [2]

Comparison of 35 and 50 {\mu}m thin HPK UFSD after neutron irradiation up to 6*10^15 neq/cm^2

Y. Zhao et al.,Comparison of 35 and 50µm thin HPK UFSD after neutron irradiation up to 6*1015 neq/cm2,Nucl. Instrum. Meth. A924(2019) 387–393, [1803.02690]. [3]ATLAScollaboration, HGTD,Technical Design Report: A High-Granularity Timing Detector for the ATLAS Phase-II Upgrade, Tech. Rep. CERN-LHCC-2020-007. ATLAS-TDR-031, CERN, Geneva, Jun, 2020

work page Pith review arXiv 2019

[3] [4]

CMS collaboration,A MIP Timing Detector for the CMS Phase-2 Upgrade, Tech. Rep. CERN-LHCC-2019-003. CMS-TDR-020, CERN, Geneva, Mar, 2019

work page 2019

[4] [5]

An LGAD-based full active target for the PIONEER experiment

Electron-Ion collider, “https://www.bnl.gov/eic/.” [6]PIONEERcollaboration, S. M. Mazza,An LGAD-Based Full Active Target for the PIONEER Experiment,Instruments5(2021) 40, [2111.05375]

work page Pith review arXiv 2021

[5] [7]

Gain suppression study on LGADs at the CENPA tandem accelerator

S. Braun et al.,Gain suppression study on LGADs at the CENPA tandem accelerator,Nucl. Instrum. Meth. A1064(2024) 169395, [2405.02550]

work page Pith review arXiv 2024

[6] [8]

M. C. J.-R. et al.,Study of ionization charge density-induced gain suppression in LGADs,Sensors22 (2022)

work page 2022

[7] [9]

Currás, M

E. Currás, M. Fernández and M. Moll,Gain reduction mechanism observed in Low Gain Avalanche Diodes,Nucl. Instrum. Meth. A1031(May, 2022) 166530

work page 2022

[8] [10]

D. S. et al.,Status of and operating experience with the university of washington superconducting booster linac,Nucl. Instrum. Meth A 287, 247(1990)

work page 1990