LAUSR

Nowadays electron channeling contrast imaging (ECCI) is widely used to characterize crystalline defects on blanket semiconductors. Its further application in the semiconductor industry is however challenged by the emerging rise of nanoscale 3D heterostructures. In this study, an angular multi-segment detector is utilized in backscatter geometry to investigate the application of ECCI to the defect analysis of 3D semiconductor structures such as III/V nano-ridges. We show that a low beam energy of 5 keV is more favorable and that the dimension of 3D structures characterized by ECCI can be scaled down to ~ 28 nm. Furthermore, the impact of device edges on the collected ECCI image is investigated and correlated with tool parameters and cross-section profiles of the 3D structures. It is found that backscattered electrons (BSE) emitted from the device edge sidewalls and generating the bright edges (edge effects), share a similar angular distribution to those emitted from the surface. We show that the collection of low angle BSEs can suppressed the edge effects, however, at the cost of losing the defect contrast. A positive stage bias suppresses edge effects by removing the inelastically backscattered electrons from the sidewalls, but low loss BSEs from the sidewalls still contribute to the ECCI micrographs. On the other hand, if segments of an angular backscatter (ABS) detector are properly aligned with the nano-ridges, BSEs emitted from the sidewall and the surface can be separated, thus leading to the completely absence of one bright edge on the surface without compromise of the defect contrast. The merging of two such ECCI images reveals the nano-ridge surface without edge effects.