New Routine to Determine Microstructures of Metallic Interconnectors Using SEM-EBSD Technique

Electron Backscatter Diffraction (EBSD) in Scanning Electron Microscope (SEM) is an ideal technology for microstructure analyses of metallic interconnectors, including phase identification, grain size and boundary characterization, and crystallographic orientation statistics. However, the application of EBSD is limited for characterizing the samples with unknown intermetallic compounds (IMCs) that are commonly formed among different interconnect materials and strongly related to failures. One reason for the limited EBSD application is that many users are not familiar with the technology that can solve the unknown phases in the current EBSDs, resulting in EBSD map data that do not provide the correct information on grain size and orientation. In this paper, we demonstrate case studies using EBSD on three typical samples of interconnectors, Al/Cu wire bonding, solder microbump, as well as through-silicon via (TSV). The software-assisted Phase Identification combining EBSD with Energy X-ray Dispersive Spectrometer (EDS) is explained as it is the key step to the sequent characterization. To avoid missing the correct phases in the IMC layers, parameter optimization based on the recently developed CMOS-based EBSD detector and the corresponding software is also presented. With the help of Phase Identification, Al2Cu and Al4Cu9 compounds both with the cubic crystal structure can be differentiated by the program in the IMC zone between Cu wire and Al pad. The further EBSD-EDS simultaneous mapping reveals that two compound layers with grain size of 50 to 400 nm are detected as the transition from the Cu wire to Al pad. In the microbump sample, multiple compounds with complex crystal structures are found, such as orthorhombic Cu3Sn and Ag3Sn, hexagonal Cu6Sn5 phase. According to the phases detected by Phase Identification, the mapping technology used to differentiate materials with very similar crystal structure, but different chemistry is applied. We can also use this technique to analyze the microstructure of TSV cross-sections across the length and at the interface. Henceforth, a new workflow of EBSD data acquisition for interconnect samples using the latest EBSD software is summarized based on the case studies.