Issue link: https://iconnect007.uberflip.com/i/992528
70 PCB007 MAGAZINE I JUNE 2018 structure, the Pd grain size of the Pd layer also seemed to small size. From these results, it was thought that Au on Cu and Pd on Au deposited as epi- taxial growth along with crys- tal orientation of under layer, and this fact resulted in bigger Pd grain on IG. WBR of IGEPIG was more superior to that of other conventional process. The superiority of WBR came from prevention of Cu and Pd diffusion to the top surface of Au from the AES results. EBSD analysis revealed that Pd grain size of IGEPIG was quite bigger than that of EPIG and Pd crystal orientation of IGEPIG was same with that of Au on Cu and Cu itself. It seemed that Au on Cu and Pd on Au were formed as epitaxial growth along with Cu crystal orientation of under layer. For IGE- PIG deposit, it was assumed that the preven- tion of Pd and Cu diffusion came from bigger Pd grain size of IGEPIG deposit and resulted in excellent WBR even if after HT. Because Pd grain size depends on the crys- tal structure of Cu, it will be necessary to in- vestigate the influence of electrolytic Cu type, grain size of Cu, the condition of etching and Au plating bath condition on Cu, and so on. Conclusions The IGEPIG process had better pattern abil- ity for narrow spaces, compared with the con- ventional ENEPIG and EPIG process because of the elimination of EN and Pd activator pro- cess. IGEPIG deposit indicated excellent WBR after HT for 16 hours at 175°C, compared with EPIG and ENEPIG deposits, even if Au thickness of IGEPIG at the top surface was same with that of EPIG and ENEPIG depos - its (Table 8). From the results of AES analysis after HT, it was confirmed that no Pd or Cu was detected at the top surface of Au when using IGEPIG de - posit, although Pd or Cu diffusion was detected when using EPIG or ENEPIG deposit. At same time, the mechanism which Pd diffusion pre - vention was relative with bigger Pd grain size was clarified by using cross sectional EBSD. Additionally, SJR of the IGEPIG deposit was al- most same with that of ENEPIG with pure Pd and was better than EPIG deposit when using Sn-3.0Ag-0.5Cu as solder ball. PCB007 References 1. Donald Gudeczauskas et al., 39 th Interna- tional Symposium on Microelectronics, Octo- ber 8–12, 2006, San Diego, California. 2. V. Vuorinen, T Laurila, H. Yu, and J. K. Ki- vilahti, J. Applied Physics, 99, 023530 (2006). 3. Yukinori Oda, Masayuki Kiso, Akira Oka- da, Kota Kitajima, Shigeo Hashimoto, George Milad, Don Gudeczauskas, 41 st International Symposium on Microelectronics, Providence, Rhode Island, Nov (2008). 4. Shigeo Hashimoto, Katsuhisa Tanabe, Ma- sayuki Kiso, Kota Kitajima, Tatsushi Someya, Don Gudeczauskas and George Milad, SMTA International Conference Proceedings, 2013. 5. Tsuyoshi Maeda, Shinsuke Wada, Katsuhi- sa Tanabe, Yukinori Oda, Shigeo Hashimoto, Don Gudeczauskas and George Milad, SMTA International, 2015. This paper was originally presented at SMTA International 2017 in Rosemont, Illinois, and published in the proceedings. Tetsuya Sasamura is researcher, final finish; Tatsushi Someya, PhD is researcher, final finish; Eriko Furuya is researcher, final finish; Katsuhisa Tanabe is department manager, final finish; Shigeo Hashimoto is senior managing director, all with C. Uyemura & Co. Table 8: Conclusions of the testing. E