Radiation Testing of Optical and Semiconductor Components for Radiation-Tolerant LED Luminaires

1. Introduction & Overview

This work, presented at the 2018 RADECS conference, addresses a critical infrastructure challenge at CERN: replacing obsolete fluorescent and sodium lighting in accelerator tunnels with modern, efficient LED technology. The primary obstacle is the harsh radiation environment, with annual levels exceeding $5 \times 10^{12}$ n_eq/cm² (1 MeV neutron equivalent in Si) and 1 kGy dose. The paper details a systematic irradiation campaign to qualify individual components—optical materials and power supply diodes—for integration into radiation-tolerant LED luminaires.

2. Components Under Test

The study focused on two critical component categories within an LED luminaire: the optical elements and the rectification diodes in the power supply.

3. Irradiation Methodology & Experimental Setup

Optical Materials: Gamma-ray irradiation was performed using a ⁶⁰Co source. The key metric for degradation was the induced Radiation-Induced Attenuation (RIA), measured spectrophotometrically. The dose rate and total integrated dose (up to 100 kGy) were carefully controlled to simulate long-term exposure in accelerator tunnels.

Semiconductor Diodes: Proton irradiation at 24 GeV/c was conducted at the CERN IRRAD facility. The primary degradation mechanism here is displacement damage, where high-energy particles knock atoms out of their lattice sites, creating defects that degrade electrical performance. Fluence levels targeted were beyond $8 \times 10^{13}$ n_eq/cm².

4. Results & Analysis

5. Key Insights & Degradation Mechanisms

6. Original Analysis: Core Insight, Logical Flow, Strengths & Flaws, Actionable Insights

Core Insight: This CERN study delivers a brutally practical truth for harsh-environment engineering: when facing ionizing radiation, material pedigree is everything, and commercial off-the-shelf (COTS) components fail in predictable, stratified ways. The real value isn't just in ranking Fused Quartz over polycarbonate, but in quantifying the performance gap under identical, realistic conditions to drive actionable component selection.

Logical Flow: The paper's structure is a model of applied research. It starts with a clear operational problem (obsolete lighting), decomposes the system into its most vulnerable sub-units (optics, power electronics), subjects representative samples to relevant stressors (gamma for optics, protons for displacement damage in semiconductors), and maps the degradation to physical mechanisms. This cause-effect chain from system need to material science is impeccable.

Strengths & Flaws: The major strength is its comparative methodology. Testing diverse materials (glasses vs. polymers) and semiconductor technologies (Si vs. SiC) side-by-side under controlled conditions provides a definitive guide. The use of high-energy protons for diode testing is also a strength, accurately simulating the mixed-field environment of an accelerator tunnel. However, a flaw is the lack of combined-effects testing. In a real luminaire, optics and electronics are irradiated simultaneously; synergistic effects (e.g., heat from diode degradation affecting plastic optics) are not explored. Furthermore, while SiC's superiority is clear, the study doesn't delve into the cost-benefit analysis, a critical factor for large-scale deployment at CERN or in nuclear facilities.

Actionable Insights: For engineers, the takeaway is unambiguous: 1) Standard plastics are a non-starter for optical elements in kGy-level fields. The search should focus on radiation-grade polymers or default to fused silica/quartz. 2) SiC is ready for prime time in power electronics for these environments. The data strongly supports its adoption over Si for rectification and switching. 3) This component-level qualification approach should be the blueprint for hardening any complex system (sensors, cameras, robotics) for use in particle accelerators, space (as supported by ESA's component testing data), or fission/fusion reactors. Don't test the whole system first; identify and ruthlessly test the weakest links.

7. Technical Details & Mathematical Models

The degradation of optical materials is often modeled by the Radiation-Induced Attenuation (RIA) coefficient:

$\alpha_{RIA}(\lambda, D) = \frac{1}{L} \ln\left(\frac{T_0(\lambda)}{T_D(\lambda)}\right)$

where $\alpha_{RIA}$ is the attenuation coefficient (cm⁻¹), $L$ is sample thickness, $T_0$ is initial transmission, $T_D$ is transmission after dose $D$, and $\lambda$ is wavelength.

For semiconductors, displacement damage is quantified by the Non-Ionizing Energy Loss (NIEL), which scales with the particle fluence $\Phi$ and a damage factor $\kappa$:

$\Delta V_F \propto \kappa \cdot \Phi$

where $\Delta V_F$ is the change in forward voltage. The damage factor $\kappa$ is significantly lower for SiC than for Si, explaining its superior hardness.

8. Experimental Results & Chart Description

Conceptual Chart: Optical Transmission vs. Dose

Imagine a chart with Total Integrated Dose (kGy, log scale) on the X-axis and Normalized Optical Transmission at 500 nm (%) on the Y-axis.

• Fused Quartz (FQ) Line: A nearly horizontal line, showing a slight decline from 100% to ~95% at 100 kGy. This indicates minimal darkening.
• Borosilicate (BS) Line: A gently sloping line, descending from 100% to around 70-80% at 100 kGy.
• PMMA & PC Lines: Two steeply plunging curves. PMMA might drop to ~30% and PC to below 20% transmission at 100 kGy, demonstrating severe failure for optical applications.

Conceptual Chart: Diode Forward Voltage Increase vs. Proton Fluence

A chart with 1 MeV n_eq Fluence (n/cm², log scale) on the X-axis and Percentage Increase in Forward Voltage ($\Delta V_F / V_{F0}$ %) on the Y-axis.

• Si Diode Line: A steep, upward-curving line, showing increases of 50%, 100%, or more at fluences above $10^{14}$ n/cm².
• SiC JBS Diode Line: A very shallow, almost linear increase, remaining below a 10-15% increase even at the highest tested fluences, highlighting its robustness.

9. Analysis Framework: A Non-Code Case Study

Scenario: A team is designing a radiation-hardened camera for monitoring inside a nuclear reactor containment building.

Application of the Framework from this Paper:

1. Decompose the System: Identify critical, radiation-sensitive sub-components: Image sensor (CMOS/CCD), protective window/lens, power regulation circuitry.
2. Define the Stressor: The environment features high gamma dose rates and neutron flux. Gamma primarily causes total ionizing dose (TID) effects, neutrons cause displacement damage.
3. Select Test Components:
   • Optics: Source samples of candidate lens materials: fused silica, radiation-resistant glass (e.g., BK7G18), and standard optical plastics.
   • Electronics: Source candidate voltage regulators: standard Si LDOs and potential SiC-based or hardened Si alternatives.
4. Execute Comparative Irradiation:
   • Irradiate all optical samples with Co-60 gamma to the expected lifetime dose (e.g., 10 kGy). Measure RIA across the sensor's spectral range.
   • Irradiate electronic components with neutrons (or high-energy protons as a proxy) to the expected fluence. Monitor key parameters like dropout voltage, noise, and quiescent current.
5. Analyze & Select: Based on data, choose the material/component with acceptable degradation. For example, the data may force the selection of a fused silica window and a specially hardened voltage regulator, while ruling out standard plastic lenses and commercial Si regulators.

This structured, component-first approach, directly inspired by the CERN paper, prevents costly failures of integrated systems by identifying show-stoppers at the material level early in the design process.

10. Future Applications & Development Directions

• Advanced Material Engineering: Development of "radiation-grade" polymers with molecular structures designed to resist color center formation, potentially using nano-composites or specific additives to scavenge radicals.
• SiC Dominance in Power Electronics: Wider adoption of SiC MOSFETs, JFETs, and JBS diodes not just in lighting but in all power conversion units within radiation environments (e.g., magnet power supplies, detector front-end power).
• Integrated Photonic Systems: Testing and hardening of optical fibers, splitters, and modulators for data transmission in accelerators and fusion reactors (e.g., ITER), where the principles of RIA are directly applicable.
• Machine Learning for Prediction: Using datasets from studies like this one to train models that predict component lifetime and degradation based on material properties and radiation spectra, accelerating the design cycle for rad-hard systems.
• Expansion to New Environments: Applying this qualification methodology to components for lunar/Martian surface applications (exposed to cosmic rays and solar particle events) and next-generation nuclear fission reactors.

11. References

1. J. D. Devine et al., "Radiation tests on LED-based lights for the LHC and other accelerator tunnels at CERN," IEEE Trans. Nucl. Sci., vol. 63, no. 2, pp. 841-847, Apr. 2016.
2. CERN Radiation Protection Group, "Calculated dose and fluence values in the LHC tunnels," CERN Internal Report, 2017.
3. A. Floriduz et al., "Radiation effects on high-power white GaN LEDs for accelerator lighting," Microelectronics Reliability, vol. 88-90, pp. 714-718, 2018.
4. M. Brugger et al., "Radiation damage studies on diodes and LEDs for the LHC and its injectors," CERN-ATS-Note-2013-004 PERF, 2013.
5. NASA Jet Propulsion Laboratory, "Silicon Carbide Electronics for Harsh Environments," [Online]. Available: https://www.jpl.nasa.gov.
6. European Space Agency (ESA), "Component Radiation Hardness Assurance Guidelines," ESCC Basic Specification No. 22900.
7. F. M. S. Lima et al., "Radiation-induced attenuation in optical fibers: A comprehensive review," IEEE Trans. Nucl. Sci., vol. 67, no. 5, pp. 912-924, 2020.
8. A. J. Lelis et al., "Basic mechanisms of threshold-voltage instability in SiC MOSFETs," IEEE Trans. Electron Devices, vol. 65, no. 1, pp. 219-225, 2018.