Issue link: https://iconnect007.uberflip.com/i/992528
68 PCB007 MAGAZINE I JUNE 2018 interface between Cu and Pd from IPF image for IGEP deposit, which Pd area had partial- ly same color with Cu area. Obviously, it in- dicated that Pd crystal orientation grew along with that of Cu for IGEP deposit. This may re- late with the crystal lattice that Cu, Au, and Pd are face-centered cubic (FCC). The gold lay- er of IGEP deposit was too thin to observe in this EBSD condition, which the sample tilt was 70 degrees and the accelerating voltage and ir- radiation current of this condition are higher than those of high resolution conditions. On the other hand, Pd grain size was very small for EP deposit which Pd layer deposited from palladium activator on copper and this crystal orientation of Pd grain was random. It was very difficult to analyze EBSD pattern be- cause the grain size was small. So, there were a lot of black parts in IPF image of EP depos- it. The average grain size of Cu as under lay- er was around 0.7 µm from the calculation by EBSD analysis, which grain boundaries was defined as more than 5 degrees. The average grain size of Pd deposit on Au layer was about 0.34 µm, which had same order with Cu grain. However, Pd grain size which deposited from Pd activator on copper was less than 0.045 µm which is an order less than that of Cu. It means the Pd layer did not grow along with Cu struc- ture. In the case of ENEPIG with pure Pd, be- cause Ni-P layer is known as an amorphous Figure 8: FIB-SEM image; (a) EPIG with 1.2 µm Pd, (b) IGEPIG with 1.2 µm Pd. Figure 9: EBSD results; (a) SEM image of IGEP, (b) IPF image of IGEP, (c) SEM image of EP on Pd activator, (d) IPF image on Pd activator, (e) color-coded map of IPF images.