Design and characterization of the POKERINO prototype for the POKER/NA64 experiment at CERN

This paper presents the design and experimental characterization of the POKERINO prototype, a PbWO$_4$-based electromagnetic calorimeter utilizing SiPM sensors, to validate its performance and energy resolution capabilities for the NA64 experiment's search for light dark matter.

The Gist

Imagine you are trying to catch a ghost. In the world of physics, this "ghost" is Dark Matter, a mysterious substance that makes up most of the universe but refuses to interact with light or normal matter in any way we can easily see.

The NA64 experiment at CERN is like a high-stakes detective agency trying to catch this ghost. They shoot a beam of tiny particles (positrons) at a thick block of material. Usually, the particles bounce off or stop, and the detectors measure exactly how much energy they had. But if a "ghost" (Dark Matter) is created during the crash, it will fly away unseen, taking some of the energy with it. The detectives look for a "missing energy" gap—a hole in the ledger where energy should be but isn't.

To do this, they need a super-accurate scale. If the scale is off by even a tiny bit, they might think the energy is missing when it's actually just a measurement error.

The Problem: The "Overloaded" Scale

The team needed a new type of scale called a calorimeter (PKR-CAL). It's made of special crystals that glow when hit by particles. To read this glow, they use sensors called SiPMs (Silicon Photomultipliers). Think of these sensors as thousands of tiny buckets waiting to catch photons (particles of light).

However, there was a big problem:

1. Too much light: When a high-energy particle hits, it creates a flood of light. If too many photons hit the sensors at once, the "buckets" get full, and the extra light spills over. This is called saturation. It's like trying to fill a bathtub with a firehose; once the tub is full, you can't measure how much more water is coming in.
2. Unstable water pressure: The beam of particles isn't perfectly steady. Sometimes it's a gentle drizzle, sometimes a sudden downpour. If the pressure changes quickly, the sensors might get confused, making the scale wobble.

The Solution: The "POKERINO" Prototype

Before building the giant, expensive final scale, the team built a small, test version called POKERINO. Think of POKERINO as a "proof-of-concept" kitchen scale before building a massive industrial weighing machine.

POKERINO is a small 3x3 grid of crystals (like a tiny tic-tac-toe board made of heavy, glowing lead-glass). Each crystal is wrapped in shiny foil to keep the light inside and read by four tiny sensors working together.

What They Did (The "Test Drive")

The team took POKERINO on a road trip to test it in three different ways:

1. The Cosmic Ray Test (The "Rain" Test): They set up POKERINO in a lab in Genoa and let natural cosmic rays (particles raining down from space) hit it. This was like testing the scale in the rain to see if it could handle random, unpredictable drops.
2. The Beam Test (The "Firehose" Test): They took it to CERN and shot a controlled beam of high-energy electrons at it. They varied the energy from 10 to 100 GeV (gigaelectronvolts). This was like testing the scale with everything from a feather to a bowling ball to see if it stayed accurate.
3. The Laser Test (The "Strobe Light" Test): They used a super-fast laser to flash light at the sensors thousands of times a second. This tested if the sensors could handle rapid-fire flashes without getting confused or losing their "gain" (sensitivity).

The Results: It Works!

The results were a huge success:

• No Spilling: Even when the light flood was huge, the team figured out a mathematical trick (a correction formula) to fix the "spillover" effect. It's like realizing your bathtub is full, but you can calculate exactly how much water would have fit if the tub were bigger.
• Steady Hands: They proved that even when the beam intensity fluctuated (like a flickering light), the sensors didn't lose their calibration. The "negative feedback" mechanism (where the sensors naturally adjust to the current flow) kept things stable.
• Precision: The scale was accurate enough to meet the strict requirements needed to catch the Dark Matter ghost. The resolution was sharp enough to spot the tiny "missing energy" signature.

The Bottom Line

The POKERINO prototype proved that the team's design works. They showed that you can use these tiny, sensitive sensors (SiPMs) in a high-energy environment where they were previously thought to be too fragile or prone to errors.

In simple terms: They built a small, test version of a super-precise scale, proved it could handle a firehose of particles without breaking, and confirmed it's ready to be used in the big experiment to hunt for the universe's biggest mystery: Dark Matter. The "ghost" hunt is officially on!

Technical Summary

1. Problem Statement and Motivation

The NA64 experiment at CERN is searching for Light Dark Matter (LDM) using a "missing-energy" technique. While the electron-beam program is well-established, a new positron-beam program (supported by the POKER ERC project) was initiated to exploit a specific LDM production mechanism: resonant $e^+e^-$ annihilation on atomic electrons.

This new approach requires a detector with extremely high energy resolution to distinguish a narrow signal peak in the missing-energy distribution. The physics requirements dictate an energy resolution of:
$$\frac{\sigma_E}{E} \simeq \frac{2.5\%}{\sqrt{E[\text{GeV}]}} \oplus 0.5\%$$
This must be achieved under demanding conditions:

• High Beam Intensity: Typical rates of 100 kHz, with instantaneous fluctuations up to 30%.
• High Energy: Up to 100 GeV positrons.
• Compactness: The detector must fit within ~50 cm along the beamline.

Standard calorimeters struggle to meet these resolution and compactness requirements simultaneously. The proposed solution is a homogeneous electromagnetic calorimeter (PKR-CAL) based on Lead Tungstate ($PbWO_4$) crystals read out by Silicon Photomultipliers (SiPMs). However, using SiPMs in high-energy physics with such high light yields poses challenges regarding saturation effects (finite pixel count) and gain fluctuations induced by beam intensity variations.

2. Methodology

To validate the PKR-CAL design before full construction, the authors developed and characterized a small-scale prototype named POKERINO.

Prototype Design (POKERINO)

• Geometry: A $3 \times 3$ matrix of $PbWO_4$ crystals ($20 \times 20 \times 220$ mm$^3$), mimicking the full $9 \times 9$ PKR-CAL design.
• Readout: Each crystal is read by an assembly of four $6 \times 6$ mm$^2$ SiPMs (Hamamatsu S14160-6010).
• Hybrid Connection: A novel hybrid circuit connects the four SiPMs per crystal. This allows them to share a bias voltage while presenting a series combination to fast pulses, minimizing readout channels and high-voltage complexity.
• Mechanical/Thermal: Crystals are wrapped in reflective (VM2000) and light-tight (Tedlar) foils. The assembly includes active water cooling to maintain stability at $20^\circ$C ($\pm 0.1^\circ$C).

Experimental Campaigns

The prototype was tested in three distinct environments:

1. Cosmic Ray Commissioning (Genova): Used the EEE telescope to verify channel functionality and establish a preliminary energy calibration using Minimum Ionizing Particles (MIPs).
2. Test Beam (CERN H6, Summer 2024):
   • Energy Calibration: Used a 120 GeV/c muon beam to calibrate the response.
   • Linearity & Saturation: Used electron beams ranging from 10 to 100 GeV to study non-linearity caused by SiPM saturation.
   • High-Frequency Response: Tested the detector's response to high-intensity hadron beams ($\pi^-$) and pulsed lasers to simulate the gain fluctuations expected at the H4 beamline (100 kHz–1 MHz).

3. Key Contributions

• First Homogeneous SiPM Calorimeter for High-Energy Physics: This work represents one of the first implementations of a fully homogeneous $PbWO_4$ calorimeter read out by SiPMs for energies in the 10–100 GeV range.
• Novel SiPM Readout Architecture: The development and validation of the "PKR-CAL-SiPM" hybrid assembly, which successfully manages bias voltage distribution and signal readout for multiple sensors per crystal.
• Saturation Correction Model: The authors derived an empirical correction function to linearize the detector response despite SiPM pixel saturation, allowing accurate energy reconstruction up to 100 GeV.
• Gain Stability Analysis: A comprehensive study of the "negative feedback" mechanism where high beam rates cause voltage drops across bias resistors, reducing SiPM gain. The authors quantified this effect and demonstrated it is manageable.

4. Key Results

• Energy Resolution:
  • The measured energy resolution fits the parameterization $\sigma_E/E = A \oplus B/\sqrt{E} \oplus C/E$.
  • Constant Term ($A$): $(0.55 \pm 0.01)\%$, meeting the design requirement of 0.5%.
  • Stochastic Term ($B$): $(1.8 \pm 0.5)\%$, consistent with the expected light yield of ~5 photoelectrons/MeV.
  • Noise Term ($C$): $(161 \pm 5)$ MeV.
• Linearity and Saturation:
  • Without correction, the detector showed clear saturation at high energies.
  • Applying an exponential correction function ($E = \alpha E_{sat} (1 - e^{-E_{true}/E_{sat}})$) restored linearity.
  • The fitted saturation parameter $E_{sat} \approx 395$ GeV corresponds to an effective cell count consistent with the SiPM specifications.
• Gain Stability (Intensity Fluctuations):
  • Theoretical modeling predicted a cutoff frequency ($f_{cut}$) of ~10 MHz for the gain drop.
  • Laser tests confirmed $f_{cut} \approx 8.3$ MHz.
  • High-intensity beam tests showed a gain reduction of only ~3.5% at maximum intensity.
  • Conclusion: The gain fluctuations induced by beam intensity variations contribute a sub-dominant term to the total energy resolution, well within the required 0.5% constant term.
• Light Yield: The system achieved an estimated light yield of ~5 photoelectrons/MeV, sufficient for the physics goals.

5. Significance

The successful characterization of the POKERINO prototype validates the technical feasibility of the PKR-CAL detector. The results confirm that:

1. SiPMs are viable for high-energy, homogeneous electromagnetic calorimetry, provided that saturation and gain fluctuations are properly modeled and corrected.
2. The energy resolution requirements for the NA64 positron program can be met, enabling the detection of the narrow missing-energy peaks required to discover Light Dark Matter.
3. The negative feedback mechanism from high beam rates does not degrade the detector performance to a level that would compromise the physics search.

This work paves the way for the construction and installation of the full-scale PKR-CAL detector at the CERN H4 beamline, marking a critical step forward in the global search for dark sector particles.