X-ray inspection of glass substrates and through-glass vias using laminography, tilted radiography, and automated defect recognition

Glass substrates with through-glass vias (TGVs) are becoming increasingly important for advanced packaging and large-area panel-level manufacturing. However, their inspection remains challenging because relevant defects may occur both in the glass volume and inside metallized vias. Typical defect classes include missing holes, residual glass, micro-cracks, voids, incomplete metallization, and delamination. While optical inspection is highly effective for surface features and high-throughput screening, it is limited once the relevant information is buried in the glass or hidden by metal filling. This paper presents an application-oriented study of X-ray inspection of glass substrates and TGVs using 2D radiography, tilted radiography (2.5D), and computed laminography (3DCL). The focus is placed on practical system parameters and their influence on defect visibility, including beam energy, magnification, imaging geometry, and detector characteristics. Special attention is given to detector efficiency and contrast sensitivity, which are critical for identifying weakly absorbing features such as residual glass or subtle density fluctuations under realistic inspection-speed constraints. Representative case studies on etched and metallized TGV structures demonstrate how different acquisition modes reveal complementary information. In addition, segmentation in Dragonfly 3D World and automated defect recognition workflows are discussed as tools for scalable defect localization and quantification. The results support a hybrid inspection strategy in which fast radiographic methods are combined with volumetric laminography and AI-assisted analysis to address the inspection needs of emerging glass-based packaging technologies.