Exploration of GHz-Scanning Acoustic Microscopy for Failure Analysis of Innovative Silicon Technologies

In this dissertation, SAM is utilized at GHz frequencies, namely the GHz-SAM, to explore its potential in detecting μm and sub-μm sized defects for various state-of-the-art FA challenges. Firstly, artefacts hindering proper analysis of the composed acoustic images were analyzed, their root cause was identified, and means to reduce their effects on the GHz-SAM images were proposed. For an in-depth understanding of the GHz-SAM images, the interaction of the acoustic waves incident at the specimen was studied utilizing a ray theory. Realizing the contribution of ‘Rayleigh waves’ was profound to the understanding of interesting features that only appear in acoustic microscopy, such as circular fringes around Through-Silicon-Vias (TSVs). Secondly, the GHz-SAM is applied in detecting defects in the Back-End-Of-Line (BEOL), where nano-indentation was used to induce on-purpose failures in BEOL stacks. The analysis of the resulting GHz-SAM images showed the capability of detecting micron-sized sub-surface defects in thin films and the potential of using the GHz-SAM in enabling the study of the mechanical resistance of the BEOL stacks to delamination. For 3D applications, the GHz-SAM was applied to bonded wafers where it showed high potential in detecting defects and in locating misalignments through metals in 3D stacks with thinned top Si wafers, which is difficult to properly detect and visualize with other FA tools. Thirdly, a numerical model was built in Python as a driving vehicle to provide deeper insight on the interaction of the focused acoustic waves with specimen and complex defects. A methodology is proposed for the analysis of the simulation results and applied on study cases of different void sizes in TSVs at different locations. The demonstrated capability of the GHz-SAM in detecting defects in innovative Si technologies, and the numerical model results, show the high potential of using the GHz-SAM as a valuable FA tool in the field of microelectronics.

Jaar van publicatie:2019

Toegankelijkheid:Closed