Low dose gamma irradiation study of ATLAS ITk MD8 diodes

✨

This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Story of the "Super-Sensitive" Silicon Chips

Imagine you are building a high-speed camera for a super-fast race car. This camera needs to take pictures in a place that is incredibly dusty, full of static electricity, and bombarded by tiny, invisible bullets (radiation). This is the job of the ATLAS ITk, a new tracking system for the Large Hadron Collider (LHC) at CERN.

The "lenses" of this camera are made of silicon diodes (tiny electronic sensors). The scientists in this paper wanted to make sure these sensors wouldn't break or go crazy when hit by radiation. Specifically, they were worried about gamma rays (a type of high-energy light) and how they affect the "leakage" of electricity in the chips.

Here is a breakdown of what they did and what they found, using simple analogies.

1. The Two Types of "Leaks"

Think of a silicon sensor like a water tank with a hole in the bottom.

Bulk Current (The Hole in the Floor): This is water leaking straight through the middle of the tank. In our silicon chips, this is caused by the radiation knocking atoms out of place inside the material.
Surface Current (The Leaky Rim): This is water trickling down the outside edge of the tank. In silicon, this happens on the very top layer where the metal touches the silicon.

The Problem: In the past, scientists knew that if you hit these chips with huge amounts of radiation, the "floor leak" (bulk) would get bigger, but the "rim leak" (surface) would eventually stop getting worse (saturate). However, they didn't know when the rim leak stopped getting worse. Was it after a little bit of radiation? Or a lot?

2. The Experiment: A "Low-Dose" Test

The researchers took special silicon chips (called MD8 diodes) and gave them a "low dose" of gamma-ray radiation.

The Analogy: Imagine you are testing a new raincoat. Instead of throwing a firehose at it immediately, you start with a gentle mist, then a light drizzle, then a steady rain. They tested doses ranging from a tiny "mist" (0.5 krad) up to a "steady rain" (100 krad).
The Special Chips: They used two types of chips:
1. Standard Chip: Just a plain sensor.
2. Guarded Chip: This one had a special "fence" (called a p-stop implant) built around the edge to stop water from trickling down the rim. This helped them measure the "rim leak" and "floor leak" separately.

3. What They Discovered

A. The "Rim" Leaks More Than the "Floor"

When they turned on the radiation, they found something surprising:

The Floor Leak (Bulk) stayed almost the same. It didn't care about this low level of radiation.
The Rim Leak (Surface) went crazy! It increased massively, even with just a tiny bit of radiation.
The Result: The total electricity "leaking" out of the chip was almost entirely because of the surface, not the inside.

The Big Mystery: They kept increasing the radiation, but the "Rim Leak" never stopped growing. It didn't hit a "ceiling" (saturation) like they thought it might.

Conclusion: The "ceiling" for surface damage must be much higher than 100 krad. It's likely somewhere between 100 krad and the massive 66 million krad the machine will eventually face. This means the surface will keep getting worse and worse as the machine runs.

B. The "Baking" Test (Annealing)

After the chips got "dirty" with radiation, the scientists tried to "clean" them by baking them in an oven.

Low Heat (60°C): Like putting a wet towel in a warm room. The leaks got slightly worse at first. The "damage" was settling in.
High Heat (Above 100°C): Like putting that towel in a hot dryer. The damage disappeared. The leaks went back to normal, almost as if the radiation never happened!
Why this matters: This tells engineers that if the chips get damaged during the machine's operation, they might be able to "heal" themselves if they get hot enough, or at least that the damage isn't permanent if you can control the temperature.

C. The Temperature Test

They also checked how the leaks changed when the chips got cold (like a winter day) or warm (like a summer day).

They found that the "leakiness" followed a predictable mathematical rule based on temperature.
The Good News: The "Floor Leak" and the "Rim Leak" reacted to temperature in the exact same way. This means the physics behind the damage is consistent, making it easier for engineers to predict how the machine will behave in the cold, dark tunnel of the LHC.

4. Why Does This Matter?

The ATLAS experiment is about to start a new, more intense phase (High-Luminosity LHC). The sensors will be working in a very harsh environment.

Before this study: Engineers were worried about the "floor" damage but weren't sure about the "rim" damage.
After this study: They know that for the initial operations (the first few years), the surface damage is the main enemy, and it will keep getting worse without stopping.
The Takeaway: This helps the engineers design the cooling systems and the electronics to handle this specific type of "leakage" so the camera doesn't get blurry or stop working.

Summary in One Sentence

The scientists tested new silicon sensors with low levels of radiation and found that the "surface" of the chips gets damaged much faster than the "inside," and unlike previous theories, this surface damage keeps getting worse without hitting a limit—at least up to the levels they tested.

1. Problem Statement

Silicon strip detectors for the Inner Tracker (ITk) of the ATLAS experiment at the High-Luminosity LHC (HL-LHC) must operate in an extreme radiation environment. While the design targets a total ionizing dose (TID) of 66 Mrad, previous studies on $\gamma$ -irradiated sensors indicated a linear increase in bulk current but a saturation of surface current at the lowest tested TID levels (66 Mrad). This created a knowledge gap: the exact TID threshold at which surface current saturation occurs was unknown. Furthermore, previous studies on ATLAS18 strip sensors could not separate surface and bulk currents due to the lack of a contactable guard ring. Understanding the behavior of these currents at low TIDs (0.5 to 100 krad) is critical for predicting detector performance during the initial operations of the new ATLAS tracker.

2. Methodology

The study utilized specific diode structures fabricated from ATLAS18 ITk production wafers by Hamamatsu Photonics:

Samples:
- MD8: Standard $n^+$ -in- $p$ diodes (0.8 cm $\times$ 0.8 cm) without a $p$ -stop implant.
- MD8p: Similar diodes but featuring a $p$ -stop implant between the bias ring and the guard ring. This design isolates the surface current flowing from the edge, allowing for precise separation of bulk and surface currents.
Irradiation: Samples were irradiated with a $^{60}$ Co $\gamma$ -source at UJP PRAHA a.s. to 14 different TID levels ranging from 0.5 to 100 krad. Irradiations were conducted in a Pb/Al charge particle equilibrium box to ensure uniform energy deposition, with samples kept below 35°C during exposure.
Measurement Protocol:
- I-V Characteristics: Measured before and after irradiation at controlled temperatures (20°C).
- Current Separation: Bulk current ( $I_{BULK}$ ) was measured through the active volume, while surface current ( $I_{SURF}$ ) was measured from the guard ring to the edge.
- Annealing Studies:
  - Isochronal: 80 minutes at 60°C.
  - Time-dependent: Up to 1280 minutes at 60°C.
  - Temperature-dependent: Steps from 60°C to 300°C.
- Temperature Dependence: Measurements performed from -50°C to +20°C to determine activation energies ( $E_A$ ).

3. Key Contributions

First Low-Dose Separation: This is the first study to successfully separate and analyze bulk and surface currents in ATLAS ITk sensors at low TIDs (sub-Mrad), filling the gap between unirradiated states and the high-dose regime (66 Mrad).
Role of $p$ -stop Implant: Demonstrated the utility of the $p$ -stop implant (MD8p design) in isolating surface leakage currents, enabling more precise radiation damage characterization.
Annealing Behavior Mapping: Provided a comprehensive map of how radiation-induced defects evolve under different annealing conditions (time and temperature) for low-dose $\gamma$ -irradiation.
Activation Energy Verification: Confirmed that the activation energy for leakage currents in $\gamma$ -irradiated silicon is consistent across bulk, surface, and total currents, even at low doses.

4. Key Results

Current Behavior vs. TID:
- Pre-irradiation: Bulk and surface currents were comparable in magnitude.
- Post-irradiation: The total current increased significantly, driven almost entirely by a rise in surface current. The bulk current remained relatively stable and showed no significant increase up to the maximum tested TID of 100 krad.
- Saturation: No saturation of surface current was observed within the 0.5–100 krad range. This implies the saturation threshold lies between 100 krad and the previously observed 66 Mrad.
Annealing Effects:
- Low-Temperature Annealing (60°C): Both bulk and surface currents showed a slight increase over time (up to ~640 minutes), suggesting the mitigation of soft breakdown at high bias voltages.
- High-Temperature Annealing (>100°C): Currents increased up to 100°C but then decreased significantly. Annealing at 300°C fully restored current levels to pre-irradiation values, indicating that $\gamma$ -induced defects in both the bulk and surface are completely annealed at high temperatures.
Temperature Dependence & Activation Energy:
- Currents followed the standard Arrhenius relationship: $I(T) = A T^2 \exp(-E_A / 2kT)$ .
- The fitted activation energies ( $E_A$ ) were consistent across total, bulk, and surface currents, averaging approximately 1.20 eV. This aligns with previous literature (Chiligarov et al.) and suggests that the dominant defect mechanism is similar regardless of the current component or specific TID (within the tested range).

5. Significance

Operational Safety: The findings confirm that at the low TIDs expected during the initial commissioning and early operation of the ATLAS ITk, the detector performance will be dominated by surface leakage currents rather than bulk damage.
Design Validation: The results validate the use of the $p$ -stop implant design for accurate sensor characterization.
Defect Management: The observation that high-temperature annealing (300°C) completely reverses low-dose $\gamma$ -damage provides a potential recovery mechanism for sensors if necessary, though the primary focus is on understanding the initial degradation.
Modeling Accuracy: By defining the linear behavior of surface currents at low doses, this study allows for more accurate simulations and predictions of detector noise and power consumption in the early phases of the HL-LHC run.