In-situ Radiation Damage Study of Silicon Carbide Detectors Subjected to Clinical Proton Beams

This study characterizes the radiation damage in Silicon Carbide (SiC) planar PiN diodes subjected to clinical proton beams, revealing a gradual compensation of effective doping concentration and establishing linear donor removal rates to predict the performance and lifetime of future radiation-hard detector technologies.

The Gist

The Big Picture: Testing "Indestructible" Detectors

Imagine you are building a super-strong, heat-resistant sensor to catch tiny, invisible particles flying through space. You want to use Silicon Carbide (SiC) instead of regular silicon because it's like the "titanium" of the semiconductor world—it handles high heat and high voltage much better.

However, there's a catch: even titanium can get tired if you hit it with enough force. This paper is about testing how much "punch" these SiC sensors can take before they start to break down. The researchers used a medical proton beam (the same kind used to treat cancer) to simulate years of radiation damage in just a few hours.

The Setup: The "Traffic Jam" Analogy

To understand what happened, let's imagine the inside of the detector as a highway.

• The Highway: The silicon carbide material.
• The Cars: The electrical charge carriers (electrons and holes) that carry the signal.
• The Traffic Lights: The doping (impurities added to the material) that keeps the traffic flowing smoothly.
• The Radiation: A storm of tiny rocks (protons) being thrown onto the highway.

The Goal: The researchers wanted to see what happens to the traffic flow when you start throwing rocks at the highway. Do the cars stop? Do the traffic lights break?

The Experiment: Two Different Approaches

The team tested sensors from two different manufacturers (CNM and onsemi) using two different methods:

1. The "Live Feed" Method (In-situ): For two samples, they set up a camera right at the beamline. They hit the sensor with a little bit of radiation, stopped, measured the traffic, hit it again, measured again, and so on. This is like watching a car crash in slow motion, step-by-step.
2. The "Aftermath" Method (Traditional): For seven other samples, they hit them all at once with a specific amount of radiation, then took them back to the lab to see the damage. This is like waiting until the crash is over to inspect the wreckage.

The Findings: What Happened to the Traffic?

1. The "Forward Current" Drop (The Roadblock)

When they tried to push electricity forward through the sensor, they noticed something strange. As the radiation dose increased, the electricity got stuck.

• Analogy: Imagine trying to drive a car forward, but a giant wall of debris keeps getting built up in front of you. You have to press the gas pedal much harder (higher voltage) just to get the car to move a little bit.
• Result: In sensors with "lighter" traffic (lower doping), this wall built up quickly. In sensors with "heavier" traffic (higher doping), the wall took longer to form.

2. The "Capacitance" Drop (The Empty Parking Lot)

Capacitance is a way of measuring how many charge carriers are available to do work.

• Analogy: Think of the sensor as a parking lot. Before radiation, the lot is full of cars (charge carriers) ready to go. As the radiation hits, it creates "potholes" (defects) that trap the cars.
• Result: The researchers saw the parking lot slowly emptying out. The cars weren't disappearing; they were getting stuck in the potholes (deep-level traps) and couldn't move. Eventually, the lot was so full of potholes that no cars could move at all. This is called "full compensation."

3. The "Donor Removal Rate" (How Fast the Potholes Form)

The most important number they calculated is the Donor Removal Rate.

• Analogy: This is the speed at which the potholes are forming. They found that for every unit of radiation, a specific number of "cars" get trapped.
• The Numbers: They found that the potholes formed at a rate between 4.2 and 6.4 per centimeter. This is a crucial number because it tells engineers exactly how long a detector will last in a high-radiation environment (like inside a particle collider).

Why Does This Matter?

You might ask, "Why do we care about sensors that get hit by rocks?"

1. Future Experiments: Scientists are building massive new particle colliders (like the FCC) that will be incredibly bright and radioactive. They need detectors that won't die after a few months.
2. Precision Timing: New detectors called LGADs (which are like super-charged versions of these sensors) rely on a very specific amount of "traffic" to work perfectly. If radiation changes the traffic flow even a little bit, the timing gets messed up, and the experiment fails.
3. Medical Tech: Since they used a medical proton beam, this research also helps improve the sensors used in cancer therapy to ensure they are measuring the dose correctly.

The Conclusion

The researchers successfully mapped out exactly how Silicon Carbide detectors degrade under low levels of radiation. They found that:

• Radiation creates "traps" that steal charge carriers.
• This happens gradually, not all at once.
• They can now predict exactly how much radiation a sensor can take before it stops working.

The Takeaway: Silicon Carbide is a tough material, but it's not invincible. By understanding exactly how it "breaks" (or rather, how it gets clogged up), scientists can design better, longer-lasting detectors for the next generation of physics experiments and medical treatments. It's like knowing exactly how many miles a car tire can last before it needs replacing, so you never get a flat tire in the middle of a race.

Technical Summary

1. Problem Statement

Silicon Carbide (SiC), specifically the 4H-polytype, is a promising wide-bandgap semiconductor for high-energy physics (HEP) detectors due to its high breakdown voltage, thermal conductivity, and radiation hardness. However, the performance of advanced SiC detectors, particularly Low-Gain Avalanche Detectors (LGADs), relies heavily on precise doping concentrations in the gain layer. Radiation-induced defects can trap charge carriers, leading to "carrier removal" (compensation), which alters the electric field and degrades detector gain and timing resolution.

While high-fluence radiation damage in SiC is well-studied, there is a lack of quantitative data regarding the early stages of radiation damage (low fluences) and the specific donor removal rates under clinical proton beams. Understanding these initial effects is critical for predicting the operational lifetime of SiC devices in future high-luminosity experiments (e.g., HL-LHC, FCC) and for optimizing LGAD designs.

2. Methodology

The study utilized planar 4H-SiC PiN diodes from two manufacturers: CNM (IMB, Spain) and onsemi (Czech Republic). The experiments were conducted at the MedAustron ion therapy center in Austria using a 252.7 MeV proton beam.

• Sample Configuration:
  • CNM Samples: Two samples (W2 and W4) with epitaxial thicknesses of 50 µm and 100 µm, respectively. W4 had a significantly higher doping concentration ($\sim 4 \times 10^{14} \text{ cm}^{-3}$) compared to W2 ($\sim 2\text{--}10 \times 10^{13} \text{ cm}^{-3}$).
  • onsemi Samples: Seven samples (PIN2–PIN9) with 50 µm epitaxial thickness and doping around $4 \times 10^{13} \text{ cm}^{-3}$.
• Irradiation Strategy:
  • Shift 1 (In-situ): Two CNM samples were irradiated in six incremental steps (up to $\sim 3.5 \times 10^{13} \text{ p}^+/\text{cm}^2$). Electrical characterizations (I-V and C-V) were performed in-situ between fluence steps using a remote-controlled switchbox and automated Python scripts. This allowed for the observation of damage evolution within the same device.
  • Shift 2 (Traditional): Seven onsemi samples were irradiated individually to specific total fluences (ranging from $1.4 \times 10^{11}$ to $6.9 \times 10^{13} \text{ p}^+/\text{cm}^2$) and characterized in a laboratory probe station post-irradiation.
• Measurements:
  • I-V: Forward and reverse current-voltage sweeps to monitor leakage and breakdown behavior.
  • C-V: Capacitance-voltage measurements (1 MHz) to extract effective doping concentration ($N_{eff}$) and depletion width.

3. Key Contributions

• In-situ Characterization: The paper demonstrates a novel methodology for performing in-situ electrical characterization of SiC detectors directly at a proton therapy synchrotron. This approach minimizes uncertainties associated with sample-to-sample variations by tracking the same device through multiple fluence steps.
• Quantification of Early-Stage Damage: The study focuses on the low-fluence regime ($10^{11} - 10^{13} \text{ cm}^{-2}$), capturing the transition from initial defect formation to full compensation, a regime often skipped in favor of high-fluence studies.
• Donor Removal Rate Determination: The authors successfully extracted linear donor removal rates ($g$) for different doping levels and manufacturers, providing a quantitative basis for radiation hardness modeling.

4. Key Results

• Forward Current Suppression:
  • In lightly doped samples (CNM W2 and onsemi), the forward current was significantly suppressed as fluence increased. The exponential rise in current was delayed to higher bias voltages, indicating the formation of an electric field barrier caused by space charges (deep acceptor-like traps).
  • In the heavily doped sample (CNM W4), this effect was negligible within the tested fluence range, suggesting higher doping requires significantly more fluence to induce similar compensation.
• Capacitance Reduction:
  • A gradual reduction in capacitance was observed before full depletion in all samples. This confirms that free charge carriers in the epitaxial layer are being trapped by radiation-induced deep-level defects.
  • In lightly doped samples, the capacitance eventually flattened (became independent of bias voltage), indicating full compensation of the epitaxial layer.
• Donor Removal Rates ($g$):
  • The effective doping concentration ($N_{eff}$) decreased linearly with fluence ($\Phi$) according to the relation $N_{eff}(\Phi) = N_0 - g\Phi$.
  • Calculated removal rates ranged from 4.2 cm⁻¹ to 6.4 cm⁻¹:
    • CNM W4: $4.48 \pm 0.26 \text{ cm}^{-1}$
    • CNM W2: $4.60 \pm 0.33 \text{ cm}^{-1}$
    • onsemi PiNs: $5.60 \pm 0.81 \text{ cm}^{-1}$
  • Extrapolation for the CNM W4 sample suggests full dopant compensation would occur at a fluence of $\sim 8.8 \times 10^{13} \text{ p}^+/\text{cm}^2$.

5. Significance

• Predicting Device Lifetime: The quantified donor removal rates are essential for modeling the operational lifetime of 4H-SiC LGADs. Since LGAD gain is highly sensitive to the doping of the multiplication layer, even small amounts of radiation-induced compensation can drastically alter performance.
• Validation of Low-Fluence Effects: The study confirms that significant radiation damage (carrier trapping and field barrier formation) occurs at fluences much lower than previously assumed necessary for total compensation, which is critical for detectors in the inner layers of future colliders.
• Feasibility of Clinical Beams: The research validates the use of clinical proton synchrotrons (like MedAustron) for controlled, low-fluence radiation damage studies, offering precise dose delivery and well-characterized beam parameters as an alternative to research reactors.
• Future Directions: The findings provide a foundation for future studies on the impact of radiation on signal generation and charge collection efficiency in SiC LGADs, guiding the development of radiation-hard detector technologies for high-luminosity physics.