Cherenkov and scintillation light separation in BGO and BSO crystals coupled to SiPMs for dual-readout electromagnetic calorimetry at future colliders

This paper demonstrates the first successful event-by-event separation of Cherenkov and scintillation light in BGO and BSO crystals coupled to SiPMs using optical filtering and waveform template fitting, validating their potential for dual-readout electromagnetic calorimetry in future $e^+e^-$ Higgs factory detectors.

The Gist

Imagine you are trying to listen to a conversation in a very loud, crowded room. You have two people talking: one is whispering a quick, sharp secret (the Cherenkov light), and the other is shouting a long, slow story (the scintillation light). Usually, the shouting drowns out the whisper, making it impossible to hear the secret.

This paper is about a team of scientists who figured out how to hear that whisper clearly, even while the shouting is going on. They did this to build a better "camera" for seeing the smallest particles in the universe.

Here is the story of how they did it, broken down into simple parts:

1. The Goal: Building a Better Particle Camera

Future particle physics experiments (like those at the "Higgs Factory") need to measure energy with incredible precision. Think of a calorimeter as a giant, high-tech bucket that catches particles. When a particle hits the bucket, it creates a splash of light.

• The Problem: Current buckets are great at measuring the "shouting" (scintillation light), but they struggle to hear the "whisper" (Cherenkov light).
• Why it matters: If you can hear both the shout and the whisper separately, you can figure out exactly what kind of particle hit the bucket and how much energy it had. This is called Dual-Readout.

2. The Ingredients: Special Crystals

The scientists used two types of special crystals, BGO and BSO.

• Think of these crystals as dense, heavy glass blocks.
• When a high-speed particle zips through them, it does two things:
  1. It makes the crystal glow brightly and slowly (Scintillation).
  2. It creates a tiny, ultra-fast "sonic boom" of light (Cherenkov), similar to the shockwave a supersonic jet makes.

The problem is that the "glow" is so bright and long-lasting that it completely hides the "sonic boom."

3. The Solution: Two Ears and a Filter

To solve this, the scientists built a special setup with two "ears" (sensors) on each end of the crystal.

• Ear #1 (The Scintillation Channel): This ear listens to everything. It has no filter. It hears the loud, long story perfectly.
• Ear #2 (The Cherenkov Channel): This ear is wearing special sunglasses (an optical filter).
  • These sunglasses only let through the very short, blue/violet wavelengths of light (the "whisper").
  • They block out most of the bright, yellow/green light (the "shout").
  • Analogy: Imagine trying to hear a flute in a rock concert. You put on noise-canceling headphones that block the drums and guitars but let the flute pass through. That's what the filter does.

4. The Magic Trick: Listening to the Rhythm

Even with the sunglasses, a little bit of the "shout" still leaks through. So, the scientists used a clever trick: Waveform Template Fitting.

• The Analogy: Imagine you are listening to a song. The "whisper" (Cherenkov) happens instantly at the very start of the beat. The "shout" (Scintillation) starts a tiny bit later and fades out slowly.
• The scientists recorded the exact shape of the sound wave. They then used a computer to act like a musical editor, separating the "fast start" from the "slow fade."
• By mathematically subtracting the slow part, they could isolate the fast part, even if they were mixed together in the same recording.

5. The Results: Success!

They tested this setup at CERN (the world's biggest particle lab) using beams of high-energy particles.

• What they found: They successfully separated the two signals!
• The Numbers: They managed to detect about 150 "whispers" (Cherenkov photons) for every billion electron-volts (GeV) of energy.
• Why this is a big deal: The goal for the future "Higgs Factory" detector was to get at least 50 whispers. They got 150! This means their method is more than good enough to build the next generation of particle detectors.

6. The Takeaway

This paper proves that we can build a "dual-readout" camera using these crystals and special sensors (called SiPMs).

• BGO crystals are like the reliable, heavy-duty workhorses with a lot of light.
• BSO crystals are like the speedy racers; they glow faster, which is great for when particles are coming in very quickly.

By combining these crystals with "sunglasses" and "sound editing," scientists can now build detectors that are sharp enough to see the secrets of the universe with unprecedented clarity. This is a crucial step toward building the massive machines needed to study the Higgs boson and other mysteries of the future.

Technical Summary

1. Problem Statement

Future particle physics experiments, specifically $e^+e^-$ Higgs factories like the proposed FCC-ee or CEPC, require detectors with unprecedented energy resolution. The IDEA detector concept proposes a hybrid calorimeter where the electromagnetic (EM) section uses homogeneous crystals. To achieve the required hadronic energy resolution ($\sim 30\%/\sqrt{E}$), the detector must employ dual-readout calorimetry. This technique involves separately measuring the scintillation light (proportional to total energy) and Cherenkov light (proportional to the electromagnetic fraction of a shower) on an event-by-event basis.

The challenge lies in separating these two light components within compact, high-density crystals (like BGO and BSO) where the scintillation signal typically overwhelms the much fainter Cherenkov signal. Previous attempts using photomultiplier tubes (PMTs) suffered from inefficient light collection. This work investigates whether modern Silicon Photomultipliers (SiPMs) combined with optical filtering and waveform analysis can successfully separate these signals in homogeneous crystals.

2. Methodology

The study utilized a dual-readout setup instrumented with Bismuth Germanate (BGO) and Bismuth Silicate (BSO) crystals, coupled to SiPMs at both ends.

• Crystal Samples:
  • BGO: High light yield ($\sim 3500$ ph/MeV), slower decay.
  • BSO: Lower light yield ($\sim 950$ ph/MeV), faster decay.
  • Dimensions: $150 \times 12 \times 12$ mm$^3$.
• Dual-Channel Readout Configuration:
  • Scintillation Channel (S): One SiPM directly coupled to the crystal end to measure total light output.
  • Cherenkov Channel (C): The opposing SiPM is separated from the crystal by a SCHOTT UG11 short-pass optical filter (transmitting $250\text{--}390$ nm). This filter suppresses the bulk of the scintillation light (emission peak $\sim 480$ nm) while allowing Cherenkov photons to pass.
• Photodetectors:
  • C-Channel: Hamamatsu S14160-6050HS ($6\times6$ mm$^2$, high PDE in UV) to maximize sensitivity to the faint Cherenkov signal.
  • S-Channel: Hamamatsu S14160-3010PS ($3\times3$ mm$^2$, high dynamic range) to handle the intense scintillation light.
• Signal Processing:
  • Waveforms were digitized at 6.25 Gs/s.
  • Template Fitting: A pulse-shape analysis was applied to disentangle the two components. The method exploits the distinct temporal profiles: Cherenkov light is prompt, while scintillation light follows a bi-exponential decay. The fitting function $f(t) = c \cdot T_c(t-t_0) + s \cdot T_s(t-t_0)$ extracts the number of Cherenkov ($c$) and scintillation ($s$) photoelectrons for each event.
• Test Beam: Experiments were conducted at the CERN SPS North Area using:
  • 120 GeV $\mu^+$ beams: To calibrate the scintillation response to Minimum Ionizing Particles (MIPs) and determine energy deposition vs. path length.
  • 10 GeV $e^+$ beams: To induce electromagnetic showers and test the dual-readout separation capabilities.

3. Key Contributions

• First Demonstration: This is the first successful demonstration of Cherenkov–scintillation separation in BGO and BSO crystals using SiPM readout.
• Novel Separation Technique: The paper validates a hybrid approach combining spectral filtering (UG11 filter) with temporal discrimination (waveform template fitting) to isolate the Cherenkov signal despite the overwhelming scintillation background.
• Calibration Framework: A robust calibration procedure was established to convert SiPM signals into photoelectrons (ph.e.) and subsequently into deposited energy, accounting for gain variations and dark counts.

4. Results

• Scintillation Response:
  • Linear correlation ($r > 0.99$) was observed between the measured light output and the simulated energy deposition for both crystals.
  • Light Yields: BGO yielded $\sim 7.0$ ph.e./MeV, and BSO yielded $\sim 2.0$ ph.e./MeV on the S-channel.
• Cherenkov Extraction:
  • The template fitting successfully separated the prompt Cherenkov component from the slower scintillation tail.
  • Angular Dependence: The Cherenkov signal showed a strong angular dependence, peaking when the Cherenkov cone was directed toward the C-channel detector.
  • Cherenkov Fraction: At the optimal angle ($120^\circ$), the Cherenkov fraction ($C/(C+S)$) reached 33% for BGO and 74% for BSO. The higher fraction in BSO is attributed to its intrinsically lower scintillation yield.
• Cherenkov Yield:
  • At $120^\circ$ (near the Cherenkov angle), the measured yields were:
    • BGO: $\sim 97$ ph.e./GeV.
    • BSO: $\sim 152$ ph.e./GeV.
  • At $180^\circ$ (parallel incidence), yields were $\sim 49$ ph.e./GeV (BGO) and $\sim 99$ ph.e./GeV (BSO).
  • The overall uncertainty is estimated at $\sim 20\%$, dominated by systematic effects in energy estimation.

5. Significance

• Validation of IDEA Concept: The results confirm that the IDEA detector concept for future Higgs factories is feasible using homogeneous crystals. The measured Cherenkov yields ($>50$ ph.e./GeV) exceed the minimum requirement needed to achieve the target hadronic energy resolution of $\sim 30\%/\sqrt{E}$.
• Material Selection:
  • BGO offers high light yield and established manufacturing but has a slower response.
  • BSO offers a faster decay time (beneficial for high-rate environments to reduce pile-up) and a higher Cherenkov fraction due to lower scintillation background.
  • The paper suggests that BGSO (Bismuth Germanate Silicate) compositions, which are intermediate between BGO and BSO, could offer an optimal balance of light yield, timing, and Cherenkov purity.
• Technological Pathway: The study proves that compact, high-granularity dual-readout calorimeters can be built using SiPMs and standard optical filters, providing a scalable solution for next-generation particle physics experiments.