Method and Error Analysis of Heavy-ion Radiation Detection Using CMOS Image Sensors

Space radiation, particularly heavy ions, poses a significant threat to the reliability and longevity of electronic systems onboard satellites. Traditional dedicated radiation detectors are often costly, bulky, and power-intensive, limiting their applicability in space missions. This study explored the feasibility of utilizing readily available complementary metal-oxide-semiconductor (CMOS) image sensors (CIS), commonly integrated into spacecraft payloads like star trackers and cameras, as a low-cost, highly compatible in-situ solution for heavy-ion radiation detection and monitoring. The primary objectives are to identify the most sensitive CIS parameters for detecting heavy-ion flux (fluence rate) and cumulative fluence, analyze the associated detection errors, and establish a robust methodology for radiation level assessment. A front-side illuminated CIS with a 4T-pinned photodiode (PPD) pixel structure was employed. Heavy-ion irradiation experiments were conducted using tantalum (Ta) ions (linear energy transfer, LET=86 MeV·cm²/mg) at the Space Environment Simulation and Research Infrastructure (SESRI). To isolate pixel array effects, readout circuits were shielded. The pixel array was strategically divided into multiple zones. Three zones were irradiated at distinct flux levels of 2×10³, 5×10³, and 1×10⁴ cm⁻²·s⁻¹ for 1 000 seconds each, achieving cumulative fluences of 2×10⁶, 5×10⁶, and 1×10⁷ cm⁻², respectively. A fourth zone served as an unirradiated control. Overlapping regions between zones provided additional fluence data points (7×10⁶ cm⁻² and 1.5×10⁷ cm⁻²). The sensor operated normally during irradiation, with images acquired remotely in real-time. Post-irradiation dark frame analysis was also performed. Key parameters extracted from the image data included dark signal (average grayscale), number of bright spots (single-ion induced transient clusters), number of bright pixels (pixels exceeding a grayscale threshold), hot pixel percentage, dark signal non-uniformity (DSNU), noise, and dark current. Linear regression analysis (minimizing least squares) determined the relationship between these parameters and irradiation flux/fluence. Detection sensitivity (S) and nonlinear error (quantified by R²) were rigorously evaluated for each parameter. The CIS demonstrates a strong capability for heavy-ion radiation detection. For flux detection, the dark signal, bright spot count, and bright pixel count all exhibit high sensitivity (＞50%) and strong linear correlations (R²＞0.98) with increasing flux. The dark signal provides the most accurate flux detection (R²=0.999 7). For cumulative fluence detection, the dark signal, DSNU, and dark current show excellent linear responses (R²＞0.97) across the tested fluence range. The dark signal again yields the highest accuracy (R²=0.995 2). While the hot pixel percentage increases with fluence, its linearity is poor. Noise shows a weak correlation with both flux and fluence. Analysis reveals that nonlinear errors are acceptably low for the optimal parameters (dark signal, bright spots for flux; dark signal, DSNU for fluence), well within the ±10% uncertainty tolerance specified in standards like GJB 548. Sensitivity remains high under varying irradiation conditions. This research successfully establishes a method for detecting heavy-ion radiation flux and cumulative fluence using standard CIS parameters. The dark signal proves exceptionally reliable for detecting both flux (R²=0.999 7) and fluence (R²=0.995 2). The number of bright spots is highly effective for flux detection (R²＞0.98), while DSNU is a sensitive indicator for cumulative fluence (R²＞0.97). The proposed method leverages existing CIS-based payloads on satellites, offering a significant advantage in cost, size, weight, and power consumption compared to dedicated radiation detectors. The low nonlinear errors meet stringent reliability assessment requirements. These findings provide a crucial experimental foundation and technical support for developing CIS-based radiation monitoring systems, enabling efficient in-orbit radiation effect assessment and online diagnosis for spacecraft electronics.