Characterization of thin optical filters for high purity Cherenkov light readout from scintillating crystals

This study characterizes thin optical filters for hybrid dual-readout calorimeters, demonstrating that while interference filters are unsuitable due to angle-dependent transmittance, specific absorptive long-pass filters effectively block over 99% of scintillation light from PWO crystals to enable high-purity Cherenkov light readout.

The Gist

Imagine you are trying to listen to a very quiet, high-pitched whistle (the Cherenkov light) while standing next to a roaring jet engine (the scintillation light). Both sounds are coming from the same source, but the jet engine is so loud it completely drowns out the whistle.

This is the exact problem physicists face when building a new type of particle detector for future particle colliders. They want to measure two different types of light emitted by crystals when hit by particles:

1. Scintillation Light: A bright, overwhelming "glow" that tells them how much energy was deposited.
2. Cherenkov Light: A faint, blue "whistle" that tells them what kind of particle it was.

To solve this, they need a special "noise-canceling" device—an optical filter—that blocks the loud roar but lets the quiet whistle through. This paper is all about testing different filters to see which one works best.

The Cast of Characters

• The Crystals (The Source): The team tested three types of heavy, dense crystals (PWO, BGO, and BSO). Think of these as the "engines" producing the light. They are like different brands of lightbulbs, each glowing with a slightly different color and intensity.
• The SiPMs (The Ears): These are tiny, super-sensitive light sensors (Silicon Photomultipliers) that act as the ears trying to hear the signal.
• The Filters (The Noise-Canceling Headphones): This is the main star of the show. The team tested two types of filters:
  • Interference Filters: These are like high-tech, thin mirrors that only reflect specific colors. They are very precise, but they are "picky" about the angle the light hits them.
  • Absorptive Filters: These are like dark, tinted sunglasses. They soak up unwanted colors of light regardless of the angle.

The Experiment: Finding the Perfect Filter

The scientists set up a lab to act as a "soundproof room" for light. They shone light through these crystals and tried to block the "roar" (scintillation) while keeping the "whistle" (Cherenkov).

1. The "Picky Mirror" Problem (Interference Filters)
The team first tried the Interference Filters. Imagine a mirror that only reflects blue light if you look at it straight on. If you tilt your head just a little, it suddenly starts reflecting red light instead.

• The Result: Because light bounces around inside the crystal and hits the filter from all different angles (like a pinball machine), these "picky mirrors" failed. They let too much of the loud roar through because the light hit them at weird angles. They were deemed unsuitable.

2. The "Tinted Sunglasses" Success (Absorptive Filters)
Next, they tried the Absorptive Filters. These are like thick, dark sunglasses that just absorb the bright glare no matter how you tilt your head.

• The Result: These worked beautifully. Specifically, a thin filter (about the thickness of a human hair, ~100 micrometers) called Kodak-24 acted like a perfect bouncer. It blocked 99% of the loud scintillation light but let the Cherenkov light pass through.

The "Ghost Light" Surprise

During the tests, they noticed something weird with one specific thick filter (Hoya-O56). It didn't just block the light; it seemed to "re-glow" with a delayed, fuzzy light.

• The Analogy: Imagine you shout into a cave, and instead of just hearing an echo, the cave walls start glowing and humming for a few seconds after you stop shouting. This "fluorescence" was a bad thing because it added confusing noise to the signal.

The Big Takeaway

The paper concludes that to build this super-precise particle detector, you shouldn't use the fancy, angle-sensitive mirrors. Instead, you should use thin, dark, absorptive filters (like the Kodak-24).

Why does this matter?
By successfully filtering out the "noise," physicists can now hear the "whistle" clearly. This allows them to distinguish between different types of particles with incredible precision, which is essential for discovering new physics and understanding the fundamental building blocks of our universe.

In short: They found the perfect pair of sunglasses that blocks the blinding glare of a jet engine so you can finally hear the quiet whistle of a particle.

Technical Summary

1. Problem Statement

The paper addresses a critical challenge in the design of a hybrid dual-readout electromagnetic calorimeter for future $e^+e^-$ colliders. The proposed detector uses high-density scintillating crystals (PWO, BGO, BSO) coupled to Silicon Photomultipliers (SiPMs) to simultaneously measure two signals:

• Scintillation (S) light: High yield, used for energy measurement.
• Cherenkov (C) light: Low yield, used to distinguish electromagnetic from hadronic showers.

The Challenge: The Cherenkov signal is orders of magnitude weaker than the scintillation signal. To achieve the required energy resolution ($30\%/\sqrt{E}$ for hadrons), the contamination of the Cherenkov channel by scintillation photons must be kept below 20%. This requires an optical filter placed in front of one SiPM to block scintillation wavelengths while transmitting Cherenkov wavelengths. The paper investigates whether existing optical filters can effectively isolate Cherenkov light from crystals that emit light over a broad angular distribution (0° to 180°).

2. Methodology

The study employed a comprehensive experimental and simulation-based approach:

• Crystal Characterization:

  • Samples: Fully polished crystals of Lead Tungstate (PWO), Bismuth Silicate (BSO), and Bismuth Germanate (BGO) with varying dimensions (lengths from 10mm to 160mm; cross-sections from $8\times8$ to $12\times12$ mm$^2$).
  • Measurements:
    • Transmittance: Measured using a Perkin Elmer Lambda 650 spectrophotometer to calculate intrinsic absorption coefficients.
    • Time Profile: Measured via Time Correlated Single Photon Counting (TCSPC) to determine decay times.
    • Light Output: Measured using a PMT and SiPMs to establish baseline photon yields (ph/MeV) and validate Geant4 ray-tracing simulations regarding light collection efficiency vs. SiPM coverage.

• Filter Characterization:

  • Types Tested: A set of ~100 $\mu$m thick filters, including Interference (Dichroic) filters (Everix) and Absorptive filters (Kodak, Hoya, Everix).
  • Testing: Transmittance curves were measured at various incidence angles (0° to 60°) to assess angular dependence.

• Performance Validation:

  • Setup: A LYSO:Ce crystal (acting as a proxy for PWO due to similar emission spectra) was read out by two SiPMs. One served as a reference (no filter), and the other ("main") was tested with various filters.
  • Source: A $^{22}$Na radioactive source (511 keV and 1275 keV $\gamma$-rays) was used to generate scintillation and Cherenkov light.
  • Analysis: The ratio of light output with and without filters was measured. Results were compared against theoretical models that convolved the crystal emission spectrum, filter transmittance, SiPM Photon Detection Efficiency (PDE), and photon angular distribution.

3. Key Contributions

• Angular Dependence Analysis: The paper definitively demonstrates that interference filters are unsuitable for this application. Their transmittance is highly sensitive to the angle of incidence. Since photons exit crystals with a broad angular distribution, interference filters fail to block scintillation light effectively at non-normal angles.
• Quantitative Modeling: Developed a robust calculation model (Eq. 4.2) that accounts for:
  • Crystal emission spectra and transmittance.
  • Filter transmittance (including angular dependence).
  • SiPM PDE wavelength dependence.
  • Cherenkov photon background contribution.
  • Geometrical light loss due to filter thickness.
• Identification of Optimal Filters: Identified specific absorptive long-pass filters (specifically Kodak-24 and Kodak-25) that meet the stringent requirements of the dual-readout calorimeter.

4. Key Results

• Crystal Properties:

  • PWO: Fast decay (4 ns), low light yield (60 ph/MeV for 130mm).
  • BSO: Intermediate decay (~76 ns), high absorption coefficient in tested samples (attributed to impurities).
  • BGO: Slow decay (~217 ns), high light yield.
  • Light output decreases by ~50% as crystal length increases from 10mm to 130mm due to absorption.

• Filter Performance:

  • Interference Filters (Everix): Showed strong angular dependence. At 30°–40° incidence, the cutoff wavelength shifted significantly, allowing scintillation light to pass. They also acted as mirrors, reflecting photons back into the crystal, increasing the reference SiPM signal.
  • Absorptive Filters (Kodak/Hoya): Showed angle-independent transmittance.
    • Kodak-24 (Cutoff ~590 nm): Blocked >99% of scintillation light from PWO crystals.
    • Kodak-29 (Cutoff ~620 nm): Blocked >97% of scintillation light from BGO/BSO crystals.
  • Fluorescence Issue: The Hoya-O56 filter exhibited delayed fluorescence (re-emission of absorbed light), which distorted the signal decay time, making it unsuitable for timing-sensitive applications.
  • Thickness Effects: Thicker filters (>1 mm) caused geometrical light loss due to refraction, reducing the Cherenkov signal. Thin filters (~100 $\mu$m) minimized this loss.

• Validation: The experimental measurements of light attenuation matched the theoretical predictions within 20%, provided that angular dependence (for interference filters) and Cherenkov background contributions were included in the model.

5. Significance

This work provides the experimental validation and selection criteria for optical filters in next-generation dual-readout calorimeters.

• Design Impact: It rules out the use of standard interference filters for high-purity Cherenkov readout in crystal calorimeters, steering the design toward thin, absorptive long-pass filters.
• Performance Guarantee: The identification of Kodak-24/25 filters confirms that it is feasible to suppress scintillation light by >99% while preserving the Cherenkov signal, satisfying the <20% contamination requirement for optimal hadronic energy resolution.
• Methodology: The study establishes a rigorous framework for characterizing optical components in high-granularity calorimetry, emphasizing the necessity of accounting for angular photon distributions in filter selection.

In conclusion, the paper successfully demonstrates that thin (~100 $\mu$m) absorptive long-pass filters with cutoff wavelengths around 590 nm are the optimal solution for isolating Cherenkov light in PWO, BGO, and BSO crystal calorimeters, enabling the realization of high-performance dual-readout detectors for future colliders.