Monolithic Active Pixel Sensors on CMOS technologies

N. Apadula, W. Armstrong, J. E. Brau, M. Breidenbach, R. Caputo, Gabriella Carinii, Alberto Collu, M. Demarteau, G. Deptuch, A. Dragone, G. Giacomini, Carl Grace, N. Graf, L. Greiner, Ryan Herbst, Gunther Haller, M. B. Jadhav, S. Joosten, C. Kenney, C. Kierans, Ji‐Hee Kim, T. Markiewicz, Yuan Mei, J. Metcalfe, Z.-e. Meziani, Tim K. Nelson, Chao Peng, Giovanni Pinaroli, P. E. Reimer, Lorenzo Rota, Marshall Scott, J. Segal, Ernst Sichterman, N. B. Sinev, A. Steinhebel, D. M. Strom, I. M. Trigger, C. Vernieri, Charles Young, M. ŻurekarXiv 2022 · DOI · arXiv

TL;DR

This paper outlines the transition of particle detector technology toward Monolithic Active Pixel Sensors (MAPS) fabricated using standard CMOS processes. By integrating sensors and readout electronics onto a single silicon chip, researchers are paving the way for massive, low-power, and ultra-precise detectors for future colliders, the Electron-Ion Collider (EIC), and space-borne gamma-ray telescopes.

What Problem Does This Paper Address?

Traditional "hybrid" silicon detectors—where the sensor and the readout chip are separate components bonded together—are reaching their limits. They are expensive to produce at scale, bulky (high "material budget"), and the interconnections between layers create parasitic capacitance that limits signal-to-noise ratios.

Future physics experiments, such as those planned for the International Linear Collider (ILC) or the Electron-Ion Collider (EIC), require hundreds of square meters of silicon. These detectors must be incredibly thin to avoid interfering with the particles they are measuring, yet precise enough to track particles with micron-scale resolution. This paper explores how MAPS technology can solve these challenges by being "monolithic" (all-in-one), allowing for thinner, cheaper, and more granular detector systems.

Methods

The authors focus on the development of MAPS using advanced CMOS technology nodes, specifically moving from 180 nm to 65 nm processes. Key technical approaches include:

  • Stitching: Since standard microchip fabrication is limited by the size of a single "reticle" (about $2.6 \times 3.2$ cm²), the authors describe "stitching" multiple reticles together on a wafer to create massive, wafer-scale sensors (up to $10 \times 10$ cm²).
  • Depleted Active Sensitive Volume (ASV): By using high-resistivity substrates and applying a bias voltage, the sensors can be fully "depleted." This allows charge to be collected via fast drift rather than slow diffusion, significantly improving radiation hardness and timing resolution.
  • Power Pulsing: For linear colliders with low duty cycles, the analog front-end is designed to shut down between bunch trains, reducing power consumption by orders of magnitude.
  • Digital Calorimetry: Instead of measuring the total charge deposited (analog), the authors propose a "digital" approach for Electromagnetic Calorimeters (ECal) that simply counts the number of activated pixels or clusters, which reduces the impact of energy fluctuations.

Datasets

As a review and status report for the Snowmass 2021 study, the paper does not use a single training dataset. Instead, it relies on:

  • GEANT4 Simulations: Extensive Monte Carlo simulations used to model detector performance, energy resolution, and particle separation (e.g., distinguishing electrons from pions).
  • Prototype Test Data: Results from early-stage chips like the AstroPix v2 and ATLASPix, which provide real-world benchmarks for energy resolution and power consumption.
  • Benchmark Requirements: Initial specifications derived from the SiD detector for the ILC and the C³ configuration.

Key Insights and Ideas

  • The "Digital" Advantage: One of the most striking insights is that a purely digital ECal (counting hits) can actually outperform traditional analog designs. By counting clusters, the detector becomes less sensitive to "Landau fluctuations" (random variations in how much energy a particle leaves behind), leading to better overall energy resolution.
  • Synergy Across Scales: The technology developed for high-energy colliders (like the 65 nm MAPS) is directly applicable to space science. The AstroPix project adapts these sensors for MeV gamma-ray detection, proving that the requirements for "low power" and "low mass" are universal across particle physics and astrophysics.
  • Overcoming the Reticle Limit: The shift toward "stitching" and "butting" chips allows for the creation of large-area trackers that were previously impossible or prohibitively expensive with hybrid technologies.

Key Results

  • Target Precision: The 65 nm MAPS prototypes aim for a spatial resolution of 7 $\mu$m and a power density of only 20 mW/cm².
  • Energy Resolution: Simulations of the digital ECal show an energy resolution of approximately $9.5\%/\sqrt{E} \oplus 0.8\%$, outperforming the analog case ($14.5\%/\sqrt{E} \oplus 1.1\%$).
  • Pion Rejection: The proposed barrel ECal for the ATHENA experiment (using AstroPix layers) demonstrates a pion rejection power of about $10^3$, which is critical for identifying rare electron signals in a sea of background noise.
  • AstroPix Performance: Early versions of the AstroPix chip achieved an energy resolution of 7% at 30 keV, meeting the stringent requirements for NASA’s AMEGO-X mission concept.

Limitations and Future Work

While promising, several hurdles remain:

  • Yield and Reliability: Creating wafer-scale stitched sensors is risky; a single defect could potentially ruin a very large and expensive chip. Developing techniques to mitigate shorts between power and ground lines is essential.
  • Power Distribution: Delivering stable power over a large, thin silicon surface without significant voltage drops is a major engineering challenge.
  • Digital Energy Resolution: While digital counting works well for Minimum Ionizing Particles (MIPs), the authors note that the digital energy resolution for non-MIP particles still requires further optimization and better readout logic.
  • Next Steps: The first small-scale 65 nm prototypes were expected in late 2022, with future work focusing on optimizing pixel sizes and exploring AI-based reconstruction algorithms to handle the "superabundant" data these sensors produce.

Summary updated

Abstract

Collider detectors have taken advantage of the resolution and accuracy of silicon detectors for at least four decades. Future colliders will need large areas of silicon sensors for low mass trackers and sampling calorimetry. Monolithic Active Pixel Sensors (MAPS), in which Si diodes and readout circuitry are combined in the same pixels, and can be fabricated in some of standard CMOS processes, are a promising technology for high-granularity and light detectors. In this paper we review 1) the requirements on MAPS for trackers and electromagnetic calorimeters (ECal) at future colliders experiments, 2) the ongoing efforts towards dedicated MAPS for the Electron-Ion Collider (EIC) at BNL, for which the EIC Silicon Consortium was already instantiated, and 3) space-born applications for MeV $γ$-ray experiments with MAPS based trackers (AstroPix).
BibTeX
@misc{apadula2022monolithic,
  author = {N. Apadula and W. Armstrong and J. E. Brau and M. Breidenbach and R. Caputo and Gabriella Carinii and Alberto Collu and M. Demarteau and G. Deptuch and A. Dragone and G. Giacomini and Carl Grace and N. Graf and L. Greiner and Ryan Herbst and Gunther Haller and M. B. Jadhav and S. Joosten and C. Kenney and C. Kierans and Ji‐Hee Kim and T. Markiewicz and Yuan Mei and J. Metcalfe and Z.-E. Meziani and Tim K. Nelson and Chao Peng and Giovanni Pinaroli and P. E. Reimer and Lorenzo Rota and Marshall Scott and J. Segal and Ernst Sichterman and N. B. Sinev and A. Steinhebel and D. M. Strom and I. M. Trigger and C. Vernieri and Charles Young and M. Żurek},
  title = {{Monolithic Active Pixel Sensors on CMOS technologies}},
  year = {2022},
  eprint = {2203.07626},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2203.07626},
}

Loading related content...

Figure