Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Surface-tension-driven self-alignment (SA) is a promising technique for heterogeneous die-level stacking. Multiple dies can be manipulated in parallel at minimal cost. A defined amount of water present between the die and a carrier substrate is used to align the components. The minimization of the water-air interface is the driving force. Most studies were performed with a completely wetted chip surface with a resulting alignment of die periphery to carrier. This achieves a maximal die-pad to carrier-pad alignment quality equal to the dicing accuracy, which is typically +/- 25 mu m. To further improve the pad-to-pad alignment accuracy, we propose a pad-assisted SA technique. With an initial placement quality of less than half a pad pitch, this technique achieves submicron SA accuracy. Furthermore, we extend the process to dies with Au studs needed in later thermo-compression bonding. Design rules to design a stable process are reported. A meniscus minimization model is built, which explains the experimental results and helps to design new SA patterns.
Edoardo Charbon, Andrea Ruffino, Yatao Peng
Edoardo Charbon, Sandro Carrara, Simone Frasca, Rebecca Camilla Leghziel