Defence of dissertation in the field of Electronics Production Technology, M.Sc. (Tech.) Antti Rautiainen
Sensors based on microelectromechanical systems have gained popularity in several applications in mobile phones, automotive, industry and medical devices. Without these sensors, self-driving cars or internet of things would not be possible to achieve. For example, modern accelerometers, gyroscopes, inclinometers and pressure sensors as well as microphones and filters have been realized with MEMS technologies. Wafer-level hermetic sealing is a very cost-effective process in manufacturing these MEMS devices. This is accomplished by bonding a processed device wafer to a cap wafer that can have also functionalities such as electrical routing for higher level integration. Due to the limitations in traditional bonding methods, several metal bonding methods have been developed intensively for wafer bonding.
Solid-liquid interdiffusion bonding (SLID) has many benefits compared to other wafer bonding methods. It enables to achieve a high re-melting temperature bond with a low temperature bonding process. Meanwhile, it has tolerance against process variation in preceding steps. However, SLID bonding has not been widely applied in industrial level due to lack of research information.
In this thesis, metallurgies for SLID bonding were studied and their reliability was evaluated with mechanical shear and tensile test as bonded and after environmental tests. Design recommendations are given, and new bond metallurgy thermodynamic information is presented. The results obtained in this thesis are directly applicable to electrical industry on various integration levels and especially in MEMS sensor wafer bonding process. For research community results provide knowledge for bond reliability and bond metallurgy development. New interconnection technologies result low stress bonds and enhance the degree of integration. Thus, manufacturing of highly miniaturized sensors is possible.
Opponents: Professor Frank Niklaus, KTH Royal Institute of Technology, Sweden and Professor Knut Aasmundtveit, University of South-Eastern Norway (USN), Norway
Custos: Professor Mervi Paulasto-Kröckel, Aalto University School of Electrical Engineering, Department of Electrical Engineering and Automation.