🔬 physics

Comprehensive study of timing resolution in plastic scintillator detectors with wavelength-shifting fiber and silicon photomultiplier readout

This paper presents a comprehensive study of timing resolution in plastic scintillator detectors using wavelength-shifting fibers and silicon photomultipliers, combining a semi-analytical model with toy Monte Carlo and full Geant4 simulations to validate predictions and generate quantitative design maps for optimizing detector performance across various materials, components, and configurations.

Original authors: Haohui Che, Guang Yang

Published 2026-07-17

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Haohui Che, Guang Yang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to catch a race car that zips past you in a blur. To know exactly when it passed, you don't just need a fast camera; you need a system that can hear the engine roar, see the headlights, and process that information instantly. In the world of particle physics, scientists build giant detectors to catch invisible particles like neutrinos or cosmic rays. These particles are so fast and small that they vanish almost instantly after hitting a detector. To figure out where they came from or what they are, scientists need to measure the exact moment they arrive with incredible precision—often down to a billionth of a second.

The detectors used for this job are often made of plastic that glows (scintillates) when a particle hits it, like a glow stick snapping in the dark. But that light doesn't travel perfectly straight to the sensor. It gets trapped in a special fiber-optic cable (a wavelength-shifting fiber) that acts like a glowing snake, carrying the light to a super-sensitive eye called a Silicon Photomultiplier (SiPM). The problem is that every step of this journey—the glowing plastic, the fiber cable, and the sensor—adds a tiny bit of "jitter" or delay. If you have too much jitter, your timing is blurry, and you can't tell the difference between two particles arriving a split second apart. Scientists have been trying to figure out exactly how to tune this whole system to get the sharpest timing possible, but it's a complex puzzle with too many moving parts to solve with just one experiment.

This paper is like a master chef's cookbook for building the perfect timing detector. The authors, H. Che and G. Yang, didn't just build one detector and test it; they built a powerful mathematical recipe that predicts exactly how well a detector will perform before it's even built. They combined a semi-analytical model (a set of smart equations), a "toy" computer simulation (a simplified game to test the rules), and a full, realistic 3D simulation (a digital twin of the real world) to map out the entire landscape of timing performance.

Here is what they found, served up with a few analogies:

The "Slow Fiber" Problem
Imagine the light from the particle is a runner. The plastic scintillator is the starting gun. The fiber is the track. The SiPM is the finish line. The authors discovered that the type of fiber you choose is the single most important factor. Some fibers, like the common Y-11 type, act like a runner who stops to tie their shoe every few meters. They absorb the light and re-emit it, but they take a long time to do it (about 7.1 nanoseconds). This "re-emission time" blurs the finish line.

The paper shows that if you switch to a faster fiber, like the YS-6 (which re-emits light in just 1.47 nanoseconds), you can cut your timing error almost in half. It's like swapping that shoe-tying runner for a sprinter. The authors calculated that this single switch is equivalent to quadrupling the amount of light you collect, which is a massive upgrade without changing the size of the detector or the electronics.

The "More Light" Myth
A common guess might be that if you just collect more light (more "photoelectrons"), your timing gets perfectly sharp, following a simple rule where doubling the light halves the error. The paper explicitly rules this out. They found that because the light signal has a "rise time" (it doesn't appear instantly at zero time but builds up), the improvement is slower than expected. Doubling the light only improves the timing by about 26–29%, not 50%. It's like trying to hear a whisper in a noisy room: adding a second person to whisper helps, but it doesn't make the room twice as quiet.

The "Mirror" Trick
What if you put a mirror at the far end of the fiber to bounce the light back? The authors tested this with their models and found it depends entirely on how long the detector is.

Short Detectors: If the detector is short (less than 20 cm), the mirror is a great idea. The light bounces back quickly enough to join the first wave of photons, giving the sensor a little boost.
Long Detectors: If the detector is long (like 100 cm or more), the mirror is useless. The light that bounces back arrives so late (about 10 nanoseconds later) that the sensor has already triggered on the first direct photon. The mirror is just a spectator that shows up after the race is over.

The "Two-Headed" Solution
For long detectors, the best strategy is to put sensors on both ends (double-ended readout). The authors confirmed that this gives a timing improvement of about 1.4 times (the square root of 2) compared to just one end. More importantly, it makes the timing uniform. With one sensor, the timing gets worse the further the particle hits from the sensor. With two sensors, the system averages the two signals, so it doesn't matter where the particle hits; the timing stays sharp everywhere.

The Electronics Bottleneck
Finally, the paper checks the "brain" of the operation: the electronics that record the time. They found that if the electronics are too slow (like the CITIROC chip used in some current experiments), they become the weak link. Even if you have the perfect fiber and tons of light, a slow clock will ruin your timing. However, once you get the electronics fast enough, they stop being a problem, and you are back to optimizing the light collection and fiber speed.

The Bottom Line
The authors validated their mathematical recipe by comparing it against thousands of simulated events and real-world data from existing experiments like SuperFGD and MINOS. The recipe matched the real data almost perfectly (within 1% error in simulations).

Their conclusion is a clear guide for anyone building these detectors:

Maximize the light: Get as many photons as possible to the sensor.
Speed up the fiber: Switching from slow fibers (Y-11) to fast ones (YS-6) is the most cost-effective upgrade you can make.
Use two sensors: For long detectors, read from both ends to keep timing consistent.
Don't worry about mirrors for long bars: They don't help.
Check your electronics: Make sure your clock is fast enough to keep up with the light.

This paper doesn't just say "we think this works"; it provides a detailed lookup table and a set of maps that allow scientists to plug in their specific detector numbers and instantly see what their timing resolution will be. It turns the art of detector building into a precise science, ensuring that the next generation of particle physics experiments can catch those fleeting particles with crystal-clear timing.

Technical Summary: Comprehensive Study of Timing Resolution in Plastic Scintillator Detectors

Problem Statement
Precise timing measurement in plastic scintillator detectors is critical for particle physics experiments, ranging from neutrino time-of-flight identification (e.g., SuperFGD in T2K) to muon detection and cosmic-ray observatories. While individual components—plastic scintillators, wavelength-shifting (WLS) fibers, silicon photomultipliers (SiPMs), and front-end electronics—are well-characterized, a generalized model connecting measurable component parameters to the achievable system timing resolution has been lacking. Existing measurements from experiments like SuperFGD, MINERvA, MINOS, and SciBar provide reference points, but they do not offer a predictive framework for optimizing new detector configurations. The challenge lies in quantifying how stochastic processes across the readout chain (scintillation emission, WLS re-emission, optical transport, single-photon detection, and electronics quantization) collectively degrade timing resolution.

Methodology
The authors develop a semi-analytical framework validated by two distinct simulation layers:

Analytical Framework: The timing resolution ( $\sigma_t$ ) is modeled as a sequence of eight functions (F1–F8) representing the photon detection chain:
- Emission & Re-emission: Scintillation is modeled as a bi-exponential rise and decay ( $f_{scint}$ ), convolved with the WLS fiber's single-exponential re-emission ( $f_{WLS}$ ).
- Detection: The combined profile is convolved with the SiPM's Single-Photon Time Resolution (SPTR), modeled as a Gaussian.
- Statistics: The timestamp is determined by the $k$ -th order statistic of $N_{pe}$ detected photoelectrons. The variance is calculated numerically using the probability density function (PDF) of the order statistic.
- Transport & Electronics: Contributions from fiber transit time spread (dependent on numerical aperture and length) and electronics quantization (TDC bin size) are added in quadrature.
- Topologies: The model accounts for single-ended, double-ended, and multi-fiber readout topologies, as well as far-end boundary conditions (absorbing, open/Fresnel, and mirrored).
Toy Monte Carlo (MC) Validation: A high-statistics Toy MC ( $2 \times 10^5$ events per point) samples the parameterized PDFs directly to validate the analytical order-statistic calculations. This covers 80 grid points across 8 fiber types and $N_{pe}$ ranges from 5 to 200.
Geant4 Optical Simulation: A full physics-based simulation tracks scintillation, WLS absorption/re-emission, fiber propagation, and SiPM detection in realistic 3D geometries. This layer validates the analytical model's ability to predict timing when evaluated at the measured $N_{pe}$ from the simulation, isolating the timing physics from light collection efficiency predictions.

Key Contributions and Results

Analytical Predictions vs. Validation: The analytical model achieves sub-percent agreement (0.9997 ± 0.0015) with the Toy MC across the parameter space. Geant4 simulations confirm the model's predictive power when calibrated with the simulated $N_{pe}$ , though absolute $N_{pe}$ prediction remains geometry-dependent.
Dominance of Photon Statistics: In over 90% of configurations, photon statistics contribute >90% of the total timing variance. The primary optimization levers are maximizing the detected photoelectron yield ( $N_{pe}$ ) and minimizing the WLS re-emission time ( $\tau_{WLS}$ ).
WLS Fiber Impact: The WLS re-emission time is the most significant hardware parameter. Switching from standard Y-11 fibers ( $\tau_{WLS} \approx 7.1$ ns) to faster fibers like YS-6 ( $\tau_{WLS} \approx 1.47$ ns) improves timing resolution by a factor of 1.7–2.2. This improvement is equivalent to increasing $N_{pe}$ by a factor of ~4, making fiber selection a highly cost-effective upgrade path.
Scaling Laws: Timing resolution follows a power law $\sigma_t \approx A \cdot N_{pe}^{-\alpha}$ with $\alpha \approx 0.44–0.49$ . This is significantly slower than the naive $N_{pe}^{-1}$ scaling expected for pure exponential emitters. The reduced scaling arises because the convolved PDF has a finite rise time, suppressing the probability of detecting photons at $t=0$ .
Readout Topology and Geometry:
- Double-ended readout provides a $\sqrt{2}$ improvement in resolution and, crucially, eliminates position dependence in long bars, offering uniform timing performance.
- Mirrored far-ends (BC3) provide modest benefits only for short detectors ( $L \lesssim 20$ cm). For long bars ( $L \gtrsim 100$ cm), reflected photons arrive too late to compete with the first direct photon, rendering the mirror effectively neutral.
Electronics Limitations: The transition from photon-statistics-limited to electronics-limited regimes occurs around a TDC bin size of $\Delta t_{TDC} \approx 2$ ns for typical configurations. Coarse electronics (e.g., CITIROC with 2.5 ns bins) can dominate the variance, masking photon-statistics improvements.
Benchmarking: The model was validated against published measurements from SuperFGD, SciBar, MINOS, and FNAL test beams. The model generally under-predicts measured resolution by 6–13% (ratios of 0.87–0.96), a discrepancy attributed to unmodeled real-world effects such as trigger time-walk, dark counts, and optical cross-talk. The FNAL strip measurement (fast fiber, high $N_{pe}$ ) showed the model correctly predicting the photon-statistics floor (0.19 ns) while the actual system resolution (0.33 ns) was dominated by readout electronics jitter.

Significance and Claims
The paper claims to provide a comprehensive, semi-analytical framework that bridges the gap between component-level parameters and system-level timing performance. Its primary significance lies in offering quantitative design maps and lookup tables for detector optimization.

The authors emphasize that the model is designed to predict timing resolution given a known or measured $N_{pe}$ , rather than predicting $N_{pe}$ from first principles (which requires geometry-specific Geant4 simulations). By separating light collection engineering from timing physics, the framework allows researchers to:

Rapidly assess the impact of switching WLS fibers or scintillator materials.
Determine the necessary $N_{pe}$ to achieve a target timing resolution.
Identify whether a specific detector configuration is limited by photon statistics, electronics, or geometry.

The work concludes that for most applications, the most effective path to improved timing is maximizing $N_{pe}$ and selecting the fastest viable WLS fiber, followed by ensuring electronics contributions remain below the photon-statistics floor. The provided lookup tables and decision matrices serve as a practical reference for optimizing plastic scintillator detectors for sub-nanosecond timing applications.

More like this