Automated quantitative analysis of void morphology evolution in AgAg direct bonding interface after accelerated aging. (November 2021)

• Record Type: Journal Article
• Authors: Yu, Z.; Xu, T.; Letz, S.; Bayer, C.F.; Schletz, A.; März, M.
• Is Part Of: Microelectronics and reliability. Volume 126(2021)
• Publication Date: 2021-11
• Subjects: Ag-Ag direct bonding -- Thin-film metallization -- Void morphology -- Image segmentation -- ImageJ; Electronic apparatus and appliances -- Reliability -- Periodicals; Miniature electronic equipment -- Periodicals; Appareils électroniques -- Fiabilité -- Périodiques; Équipement électronique miniaturisé -- Périodiques; Electronic apparatus and appliances -- Reliability; Miniature electronic equipment; Periodicals; 621.3815
• DOI: 10.1016/j.microrel.2021.114285
• Languages: English
• ISSNs: 0026-2714

Abstract

Abstract: Silver-to-silver direct bonding has been developed as an attractive joining technique for 3D power integration which can form high-strength and low-resistance Ag joints under solid-state conditions. The bonding process starts with the initial contact and plastic deformation of the surface asperities, resulting in a series of voids between these surface tips. To reliably assess the state of the direct bonded joints, precise characterization and quantification of the microstructure are required. Moreover, the key to understanding the evolution of this AgAg bond during reliability testing is the microstructural characterization of the interfacial voids in a statistically relevant manner. In this work, the void identification and analytical characterization of direct bonded joints are implemented by an automated method based on image segmentation. The current experimental results show that with this method, it is possible to track and statistically analyze the void morphology evolution during accelerated aging. This provides new insights into the fundamental void shrinkage mechanisms, which can be further used to improve the manufacturing process and lifetime and reliability of the AgAg joints. Highlights: Automated SEM image analysis based on machine learning Quantitative analysis of the morphological evolution of the nano-scale voids in diffusion bonded silver joints Visualization of the void statistics during lifetime tests