Development of Pixelated Capacitive-Coupled LGAD (ACLGADpix) Detectors

This paper presents the development and characterization of pixelated Capacitive-Coupled LGAD (ACLGADpix) detectors with a 100 $\mu$m pitch, demonstrating uniform high-precision timing performance through tests with beta rays, infrared lasers, and electron beams for future high-luminosity collider applications.

The Gist

The "Super-Speedy Camera" for the Subatomic World

Imagine you are trying to take a photo of a chaotic street festival. It's crowded, people are moving fast, and everyone is wearing the same outfit. If you just take a normal photo, you'll get a blurry mess where you can't tell who is who or where they are going.

Now, imagine if you could take a photo so fast that you could freeze every single person in mid-step, and you could also tell exactly when they took that step. That is essentially what particle physicists are trying to do at giant machines like the Large Hadron Collider (LHC). They need to track subatomic particles that are crashing into each other millions of times per second.

This paper describes a breakthrough in building a new kind of "camera sensor" that can do exactly that: track particles with incredible speed and precision.

The Problem: The "Dead Space" Dilemma

For years, scientists used a type of sensor called an LGAD (Low-Gain Avalanche Diode). Think of an LGAD like a highly sensitive microphone that can hear the "click" of a passing particle and tell you exactly when it happened (within 20 trillionths of a second!).

However, to make these microphones small enough to track individual particles in a crowded crowd, scientists had to cut them into tiny squares (pixels). The problem? To cut them, they had to put "fences" (insulation) between the squares. These fences took up space, creating "dead zones" where no sound could be heard. As the squares got smaller, the fences took up more space than the microphones themselves, making the camera useless for fine details.

The Solution: The "Floating Floor" (AC-LGAD)

The team in this paper (from KEK and the University of Tsukuba in Japan) came up with a clever workaround. Instead of cutting the microphone into separate pieces, they made one giant, continuous microphone floor.

Then, they placed a "floating floor" (a layer of metal electrodes) on top of it, separated by a thin insulating sheet (like a piece of plastic).

• The Analogy: Imagine a large, continuous trampoline (the sensor). You don't cut the trampoline; you just place a grid of trampolines above it. When someone jumps on the main trampoline, the vibration travels up through the air (capacitive coupling) to the specific grid square above them.
• The Benefit: Because the main trampoline is continuous, there are no "dead zones" or fences. Every inch of the sensor is active. This is called AC-LGAD.

The Experiment: Testing the New Sensor

The researchers built a prototype of this sensor with tiny pixels (100 micrometers wide—about the thickness of a human hair). They tested it in three ways:

1. The Beta-Ray Test (The Stopwatch): They shot radioactive particles at the sensor.

   • Result: It was incredibly fast. It could time the arrival of a particle with a precision of 25.3 picoseconds. To put that in perspective: if a picosecond were a second, a second would be 31,000 years. This proves the sensor is fast enough for the most demanding future experiments.

2. The Electron Beam Test (The Obstacle Course): They fired a beam of electrons at the sensor to see if it could track where the particles went.

   • Result: It caught the particles 99% of the time. Crucially, it didn't miss any particles near the edges of the pixels. This confirms that their "no dead zone" design actually works.
   • Spatial Resolution: It could pinpoint the location of a particle within about 24 micrometers. That's like being able to tell which specific grain of sand a fly landed on in a beach.

3. The "Crosstalk" Test (The Whisper Test): They wanted to make sure that when a particle hit one pixel, it didn't accidentally make the neighboring pixels think something hit them too.

   • Result: The signal stayed very focused. If a particle hit the center of a pixel, that pixel did almost all the talking. The neighbors stayed quiet. This is vital so the computer doesn't get confused about where the particle actually was.

Why Does This Matter?

The future of particle physics (like the High-Luminosity LHC) will be incredibly crowded. There will be so many particle collisions happening at once that old sensors will get confused, like trying to listen to one conversation in a stadium full of screaming fans.

This new Pixelated AC-LGAD sensor is the solution. It combines:

• Super Speed: It can separate events that happen almost simultaneously.
• Perfect Coverage: No dead zones, so it catches everything.
• Fine Detail: It can see exactly where a particle is.

By adding this "time" dimension to the usual "space" dimension, scientists can build 4D tracking detectors. This will allow them to untangle the mess of future collisions, helping us understand the fundamental building blocks of the universe with unprecedented clarity.

In short: They figured out how to make a sensor that is both incredibly fast and perfectly detailed, without leaving any blind spots, paving the way for the next generation of discovery in physics.

Technical Summary

1. Problem Statement

Future high-energy collider experiments, such as the High-Luminosity Large Hadron Collider (HL-LHC) and next-generation hadron colliders, will face unprecedented track densities due to severe pileup conditions (140–200 interactions per bunch crossing at HL-LHC, potentially >1000 in future colliders).

• The Limitation: Conventional tracking detectors relying solely on spatial information suffer from ambiguities in track reconstruction under these conditions. While "4D tracking" (combining spatial and precise timing information, $\sim$10–20 ps) is the solution, current technologies face a trade-off.
• The Specific Challenge: Standard Low-Gain Avalanche Diodes (LGADs) offer excellent timing resolution ($\sim$30 ps) but require junction termination extensions and isolation structures between pixels. As pixel pitch decreases, these "dead" inactive regions dominate the sensor area, making conventional segmented LGADs unsuitable for fine-pitch pixel tracking.

2. Methodology

The authors developed and tested Pixelated AC-Coupled LGADs (ACLGADpix) to overcome the dead-area limitation while maintaining high timing and spatial resolution.

• Device Architecture:
  • Unlike conventional LGADs with segmented gain layers, the ACLGAD features a continuous gain layer across the entire active area.
  • Signal readout is achieved via capacitive coupling through a dielectric layer to segmented metal electrodes (pixels).
  • This design preserves a 100% fill factor, eliminating dead regions between pixels.
  • Signal sharing between neighboring pixels is controlled by the balance between coupling capacitance and the sheet resistance of the resistive $n^+$ layer.
• Sensor Specifications:
  • Pixel Pitch: $100\,\mu\text{m} \times 100\,\mu\text{m}$.
  • Active Thickness: $20\,\mu\text{m}$ (thin layer reduces timing jitter caused by charge deposition fluctuations).
  • Fabrication: Manufactured by Hamamatsu Photonics K.K. in collaboration with KEK and the University of Tsukuba.
• Experimental Setup:
  • Beta-Ray Test: Used a $^{90}\text{Sr}$ source with a Microchannel-Plate Photomultiplier Tube (MCP-PMT) as a timing reference ($\sigma_t \approx 10$ ps) to measure intrinsic sensor performance.
  • Electron Beam Test: Conducted at the KEK PF-AR test beam line using a 3 GeV electron beam. The setup included a tracking telescope (FE-I4 and MALTA2 pixel detectors) for precise track reconstruction and an MCP-PMT for timing.

3. Key Contributions

• First Comprehensive Study: This paper presents the first full experimental characterization of pixelated AC-LGAD sensors with a $100\,\mu\text{m}$ pitch, bridging the gap between the sensor concept and a realistic 4D pixel detector.
• Validation of Full Fill Factor: Demonstrated that the AC-coupled architecture successfully eliminates the efficiency loss at pixel boundaries typically seen in segmented LGADs.
• Performance Benchmarking: Provided critical data on timing resolution, detection efficiency, spatial resolution, and crosstalk for a fine-pitch LGAD sensor.

4. Key Results

A. Timing Resolution

• Beta-Ray Measurement: Achieved a timing resolution of $\sigma_t = 25.3 \pm 0.1$ ps at a bias voltage of 105 V.
  • Significance: This performance is comparable to high-performance pad-type LGADs, proving that fine pixelation ($100\,\mu\text{m}$) does not significantly degrade timing capabilities.
• Electron Beam Measurement: Observed a resolution of 40–45 ps.
  • Analysis: The degradation is attributed to experimental conditions (multiple scattering at 3 GeV and system-level contributions) rather than the sensor itself.

B. Detection Efficiency

• Achieved an overall detection efficiency of $99.0 \pm 0.3\%$ across the active area.
• Uniformity: No significant efficiency loss was observed near pixel boundaries. This confirms the "full fill factor" design, contrasting sharply with conventional segmented LGADs which suffer from dead zones.

C. Spatial Resolution

• Intrinsic spatial resolution was determined by subtracting the tracking telescope resolution from the residual distribution:
  • $\sigma_x = 23.8 \pm 4.7\,\mu\text{m}$
  • $\sigma_y = 24.9 \pm 4.0\,\mu\text{m}$
• These values align with theoretical expectations for a $100\,\mu\text{m}$ pitch detector using binary readout.

D. Crosstalk

• Evaluated via the pulse-height ratio ($R_i$) across seven simultaneously read channels.
• Result: For hits near the pixel center, the signal is almost entirely confined to a single channel ($R_i \approx 1$). Crosstalk is localized strictly to the pixel boundaries, indicating excellent channel separation suitable for high-occupancy environments.

5. Significance

This work validates the AC-LGAD concept as a viable solution for future 4D tracking detectors. By successfully combining:

1. Excellent Timing: $\sim$25 ps resolution.
2. High Efficiency: $\sim$99% with no dead zones.
3. Fine Granularity: $100\,\mu\text{m}$ pitch with $\sim$24 $\mu$m spatial resolution.
4. Controlled Crosstalk:

The authors demonstrate that pixelated AC-LGADs can meet the rigorous requirements of future collider experiments (like the HL-LHC) where separating tracks in both space and time is critical for physics discovery. This technology represents a major step forward in semiconductor detector development for high-luminosity physics.