Precision timing at the HL-LHC with the CMS MIP Timing Detector: current progress on validation and production

This paper presents the design, physics motivation, and current production and development status of the CMS MIP Timing Detector, a new 4D timing layer designed to mitigate high-luminosity pileup challenges through its Barrel and Endcap subsystems.

The Gist

Imagine the Large Hadron Collider (LHC) as the world's most powerful particle accelerator, a giant ring where scientists smash protons together to recreate the conditions of the early universe. For years, it has been a discovery machine. But soon, it's getting a massive upgrade called the High-Luminosity LHC (HL-LHC).

Think of the current LHC as a busy highway with about 50 cars passing a specific point at the same time. The upgrade will turn that highway into a gridlock nightmare with 200 cars all crashing into each other at once. This "traffic jam" of particles is called pileup.

The problem? When 200 collisions happen at once, it becomes incredibly hard for the detectors to figure out which particle came from which crash. It's like trying to find a specific conversation in a stadium full of 200 screaming fans all talking at once.

To solve this, the CMS experiment (one of the big detectors at the LHC) is building a new "super-sensor" called the MIP Timing Detector (MTD). Here is how it works, broken down into simple terms:

The Super-Sensor: A 4D Camera

Currently, the CMS detector takes 3D photos of particle collisions (height, width, depth). The new MTD adds a fourth dimension: Time.

Imagine you are at a party where everyone is clapping. If you just look at the room, you see a blur of hands. But if you have a camera that can see exactly when each clap happens (down to a trillionth of a second), you can separate the claps. Even if two people clap at the exact same spot, if one claps a tiny fraction of a second later, you know they are different events.

The MTD can measure time with a precision of 30 to 60 picoseconds. To put that in perspective:

• A picosecond is to a second what a second is to 31.7 years.
• This speed is so fast that it can separate particles that are born in the same millimeter but at slightly different times.

The Two Halves of the Detector

Because the particle storm hits the detector differently in the middle versus the edges, the MTD is built with two different "tools" for two different jobs.

1. The Barrel (The Middle): The Crystal Flashlight

The middle part of the detector (the Barrel Timing Layer or BTL) is like a giant, hollow cylinder.

• The Material: It uses special crystals (LYSO:Ce) that glow when hit by a particle, similar to how a glow stick lights up when you snap it.
• The Reader: At both ends of these crystal bars, there are tiny light sensors called SiPMs (Silicon Photomultipliers). They act like super-sensitive eyes that catch the glow.
• The Challenge: These crystals are in a very harsh environment. Over time, the radiation will "tired" the sensors, making them noisy (like static on a radio).
• The Fix: The team put tiny refrigerators (Thermoelectric Coolers) on the sensors to keep them freezing cold (-45°C). Cold sensors are quieter. They also have a "heating cycle" (like a self-cleaning oven) to reset the sensors if they get too damaged.
• Status: This part is already being built in factories around the world and will be installed by 2026.

2. The Endcaps (The Ends): The Avalanche Sensors

The ends of the detector (the Endcap Timing Layer or ETL) face a much more intense radiation storm—about 30 times worse than the middle.

• The Material: Instead of crystals, this part uses LGADs (Low Gain Avalanche Diodes). Think of these as tiny, self-amplifying microphones. When a particle hits them, they don't just detect it; they create a small "avalanche" of electrons to make the signal louder and faster.
• The Challenge: The radiation is so strong it can break the internal structure of these sensors.
• The Fix: Engineers tweaked the chemical makeup of the sensors (adding carbon) to make them tougher. They also carefully control the electricity (voltage) to keep them working without burning out.
• Status: The design is finished, and they are currently testing the final prototypes. This part will be installed later, around 2029.

Why Does This Matter?

With this new 4D timing system, the CMS detector will be able to:

1. Unscramble the Mess: It will separate the 200 overlapping collisions, allowing scientists to see the "real" event clearly.
2. Find Rare Treasures: It will help find very rare particles (like the "Higgs boson" twins) that are currently hidden in the noise.
3. Catch Ghosts: It can help spot "long-lived particles" that travel a bit before decaying, which could be clues to Dark Matter or new physics beyond our current understanding.

The Bottom Line

The CMS collaboration is essentially giving the LHC a pair of "super-glasses" that can see time. By the time the High-Luminosity LHC starts its full run in the 2030s, this new detector will allow physicists to look through the chaos of 200 simultaneous collisions and find the hidden secrets of the universe with unprecedented clarity.

The Barrel is currently being assembled (like building the frame of a house), and the Endcaps are in the final testing phase (like wiring the electricity). Once installed, they will revolutionize how we see the subatomic world.

Technical Summary

1. Problem Statement

The High-Luminosity Large Hadron Collider (HL-LHC), scheduled to begin operations in 2030, will increase the integrated luminosity to 3000 fb⁻¹. This upgrade introduces extreme operating conditions characterized by a significant rise in pileup (simultaneous proton-proton collisions per bunch crossing), increasing from the current $\mathcal{O}(50)$ to $\mathcal{O}(200)$.

• Challenge: High pileup degrades event reconstruction, making it difficult to associate tracks with their correct primary vertices and identify physics objects (e.g., leptons, jets, missing transverse energy).
• Limitation of Current Systems: Standard 3D spatial reconstruction is insufficient to disentangle overlapping vertices that occur at the same spatial location but at different times. Collision vertices have a time spread of approximately 200 ps, which current detectors cannot resolve with sufficient precision.
• Goal: To mitigate pileup effects and maintain high physics sensitivity, the CMS experiment requires a new detector capable of measuring the arrival time of charged particles with a precision of 30–60 ps, enabling 4D vertexing (space + time).

2. Methodology and Detector Design

The solution is the MIP Timing Detector (MTD), a new timing layer integrated between the tracker and calorimeters, covering the pseudorapidity range $|\eta| < 3$. Due to differing radiation environments and mechanical constraints, the MTD is divided into two subsystems employing distinct technologies:

A. Barrel Timing Layer (BTL)

• Coverage: $|\eta| < 1.48$ (Surface area: 38 m²).
• Technology: LYSO:Ce (Lutetium Yttrium Oxyorthosilicate) Scintillating Crystals read out by Silicon Photomultipliers (SiPMs).
• Design:
  • Composed of crystal bars ($3.12 \times 3.75 \times 54.7$ mm³) read at both ends to provide two independent time measurements.
  • Total of 331,776 readout channels.
  • Radiation Mitigation: To combat radiation-induced Dark Count Rate (DCR) in SiPMs (up to 20 GHz per SiPM), the system employs:
    • Reduced SiPM overvoltage (from 3.5 V to ~1 V).
    • Dedicated noise filtering in the TOFHIR readout ASIC.
    • Thermoelectric Coolers (TECs) maintaining SiPMs at -45°C (reducing DCR by 10x) with periodic annealing at 60°C.
  • Optimization: 3.75 mm sensor thickness and 25 µm SiPM cell size were selected to maximize light output while minimizing noise.

B. Endcap Timing Layer (ETL)

• Coverage: $1.48 < |\eta| < 3.0$ (Active area: 14 m²).
• Technology: Low Gain Avalanche Detectors (LGADs).
• Design:
  • LGAD sensors are 50 µm thick to minimize Landau fluctuations and provide fast signal rise times.
  • Gain layer optimized for a gain of 10–30.
  • Radiation Mitigation: The inner radius faces fluences up to $1.6 \times 10^{15}$ neq/cm² (30x higher than BTL). Mitigation strategies include:
    • Carbon co-implantation to reduce acceptor removal (gain degradation).
    • Operation at bias voltages where the bulk electric field $E_{bulk} < 11.5$ V/µm to avoid single-event burnout.
  • Readout: Custom ETROC (Endcap Timing Read-Out Chip) ASIC, low-power (~4 mW/channel), designed for harsh radiation environments.
  • Total of ~8 million readout channels.

3. Key Contributions and Validation

The paper reports on the transition from R&D to mass production, highlighting successful validation of both subsystems.

BTL Validation and Production Status

• Test Beam Results: Modules were tested with various irradiation levels ($1\times10^{13}$ to $2\times10^{14}$ neq/cm²).
  • Non-irradiated: Achieved ~25 ps time resolution.
  • End-of-Life (Irradiated): Projected resolution degrades to ~55 ps, meeting the design target.
  • Performance breakdown confirmed that photo-statistics, electronic noise, and DCR are within expected limits.
• Production Progress:
  • 100% of LYSO crystals and SiPMs are produced and qualified.
  • ~80% of Sensor Modules (SM) assembled.

  • 50% of Detector Modules (DM) produced.

  • ~10% of Readout Units (Trays) assembled and tested.
  • Production is distributed across four centers (Beijing, Caltech, Milano, UVA) with unified QA/QC. Completion expected by February 2026.

ETL Validation and Production Status

• Performance Validation: Prototypes combining LGAD sensors and ETROC chips demonstrated the target time resolution across a bias range of ~70 V.
• Design Finalization: Sensor thickness (50 µm), pitch (1.3 mm), and array size (16×16) are finalized.
• Production Plan: ~8,000 modules to be assembled across four sites (Fermilab, Boston University, INFN, IFCA). Installation is scheduled for mid-2029.

4. Results

• Time Resolution: Both subsystems meet the stringent requirement of 30–60 ps.
  • BTL: 25 ps (start) $\to$ 55 ps (end).
  • ETL: Target 35 ps (start) $\to$ <50 ps (end).
• System Integration: Full acquisition chains for both BTL and ETL have been tested and confirmed to function correctly.
• Scalability: The mass production infrastructure is operational, with the first BTL tray assembled in March 2025 (per the paper's timeline), signaling the start of full-scale deployment.

5. Significance

The successful implementation of the MTD represents a paradigm shift for the CMS experiment:

1. 4D Vertexing: By separating vertices in time, the MTD will drastically reduce pileup contamination, allowing for the reconstruction of complex events that are currently indistinguishable.
2. Physics Sensitivity: Improved object identification (b-tagging, lepton isolation, MET) will enhance sensitivity to rare processes.
   • Impact: In multi-object final states (e.g., di-Higgs searches), the timing gain is equivalent to 2–3 additional years of data taking.
3. New Capabilities: The detector enables Time-of-Flight (ToF) measurements, improving particle identification and sensitivity to Beyond Standard Model (BSM) physics, particularly Long-Lived Particles (LLPs).
4. Technological Milestone: The paper validates the feasibility of using LYSO/SiPM and LGAD technologies in high-radiation, high-rate environments, setting a precedent for future high-luminosity collider upgrades.