H. Villarraga-Gómez
ZEISS,
United States
Keywords: failure analysis, non-destructive testing, 3D X-ray microscopy, electronics, deep learning
Summary:
This presentation introduces advanced workflows that integrate X-ray microscopy (XRM), nanoscale tomography, and deep learning (DL) to visualize the interiors of electronic devices and assemblies. These techniques facilitate the study of internal components for failure analysis (FA). Newly developed methods are introduced, including DL-based algorithms for 3D image reconstruction, which enhance scan quality by improving contrast and reducing noise. A key feature of this presentation is the introduction of a DL-based tool called DeepScout. This tool utilizes 3D XRM scans from specific regions of interest as training data to upscale high-resolution images from low-resolution datasets, effectively broadening the field of view using a neural network model. These workflows can function independently or complement other multiscale correlative microscopy evaluations, such as electron microscopy. They provide valuable insights into the inner workings of electronic packages and integrated circuits, spanning multiple length scales from macroscopic features (hundreds of mm) to microscopic details (tens of nm). Understanding advanced electronic systems through X-ray imaging and machine learning, possibly supplemented by additional correlative microscopy investigations, can accelerate development, enhance cost efficiency, and simplify FA and quality inspections of printed circuit boards (PCBs) and devices assembled with emerging technologies. Key Points to Cover 1. X-ray Technologies Overview o Initially developed for medicine, X-ray technologies are now widely used in industrial applications for analyzing electronic devices. 2. Non-Destructive Testing (NDT) Techniques o Key methods include Radiography, Laminography, and Computed Tomography, all crucial for detecting flaws in electronic components. 3. Inspection of Electronic Packages o Various packages, such as hybrids, power devices, and ultra-fine pitch devices, are inspected to identify voids, solder defects, and alignment issues. 4. Challenges with Advanced Packages o Traditional methods face limitations in depth resolution and material overlap, necessitating advanced technologies like CT and 3D XRM for improved analysis. 5. Growth of the Electronics Industry o The increasing demand for high safety and quality standards highlights the importance of advanced NDT in FA to meet these requirements. 6. Modern 3D X-ray Approaches o The integration of DL algorithms enhances imaging and analysis capabilities for PCBs and electronic devices. 7. Integration with Other Imaging Techniques o Combining X-ray methods with electron microscopy enhances inspection capabilities, allowing for more comprehensive evaluations. Key Learnings 1. Advanced Workflows o Understanding how workflows that combine XRM, nanoscale tomography, and DL enhance visualization of electronic devices. 2. Failure Analysis (FA) o Gaining insights into the internal components of electronic assemblies to aid in FA. 3. Deep Learning Integration o Exploring DL algorithms for improved 3D image reconstruction, focusing on scan quality enhancement. 4. DeepScout Tool o Learning about DeepScout, which utilizes 3D XRM scans to upscale low-resolution datasets for high-resolution insights. 5. Complementary Techniques o Understanding how these workflows can operate independently or alongside other microscopy techniques. 6. Multiscale Insights o Analyzing electronic packages and integrated circuits across various length scales. 7. Benefits of Advanced Imaging o Recognizing the potential of these techniques to accelerate development, improve cost efficiency, and simplify FA and quality inspections for PCBs and new technologies.