Speaker
Description
The development of high-performance X-ray and gamma-ray detectors is fundamentally constrained by the need to balance detection efficiency, signal stability, material thickness, and long-term radiation tolerance.[1–3] Despite extensive experimental efforts, the existing literature remains fragmented across material classes and device architectures, limiting the establishment of generalizable design principles. In this work, machine learning is employed as an interpretative framework to integrate the experimental data on radiation-sensitive materials and extract physically meaningful performance trends. A curated dataset was constructed from experimental studies including detection efficiency, luminescence yield, response time, radiation-induced degradation, and device architecture under relevant irradiation conditions. Normalized sensitivity and stability indices were derived to enable cross-study comparison. Dimensionality reduction, feature-importance analysis, and unsupervised clustering were applied to map the detector design space. The analysis reveals that the primary performance trade-off is captured by a dominant principal component explaining approximately 55–60% of the variance, corresponding to the balance between signal yield and material thickness, while a second component (20–25%) reflects radiation stability and operational durability. Organic, inorganic, hybrid, and multilayer detector systems occupy distinct performance regimes. Multilayer and hybrid architectures consistently exhibit 15–25% higher thickness normalized sensitivity at comparable stability levels. Feature-importance analysis identifies effective density, interfacial configuration, and emission efficiency as the dominant controlling parameters. These results demonstrate that radiation detection performance is primarily governed by architecture-enabled optimization rather than material composition alone. The proposed machine-learning-driven framework provides quantitative and physically meaningful design guidance for the development of compact, sensitive, and durable X-ray and gamma-ray detectors for potential applications.
Keywords: X-ray detectors; Gamma-ray detection; Machine learning; Hybrid materials; Detector architecture; Radiation stability
References
[1] B. Hou, Q. Chen, L. Yi, P. Sellin, H.-T. Sun, L.J. Wong, X. Liu, Materials innovation and electrical engineering in X-ray detection, Nat Rev Electr Eng 1 (2024) 639–655. https://doi.org/10.1038/s44287-024-00086-x.
[2] M. Chen, C. Wang, W. Hu, Organic photoelectric materials for X-ray and gamma ray detection: mechanism, material preparation and application, J. Mater. Chem. C 9 (2021) 4709–4729. https://doi.org/10.1039/D1TC00525A.
[3] Z. Lin, S. Lv, Z. Yang, J. Qiu, S. Zhou, Structured Scintillators for Efficient Radiation Detection, Advanced Science 9 (2022) 2102439. https://doi.org/10.1002/advs.202102439.