Issue link: https://iconnect007.uberflip.com/i/270052
66 SMT Magazine • March 2014 In these cross-sections there are not re- ally any differences to the interface between SnAg3.5 and HT1. The HT1 is finely dispersed, and offers more homogeny, and the SnAg 3 shows Ag 3 Sn intermetallic. The Innolot already shows some defects between the substrate and the solder joint inside the interface. Another effect with the Innolot was some die loss after the TCT, which means there was a complete crack inside the assembly. Based on that result, a product qualification with SnAg3.5 and HT1 was made. Product Qualification for hT1 The current requirements for such DCB product qualification are N=1000 at -04/+125°C with 30'/10''/30' without any electrical de- fects. Therefore, the passive test was made with these requirements and the electrical analyses as well as some cross-sections after N=250/500/750/1000/1500/2000, plus end of life (EOL). Both alloys fulfill this technical re- quirement, which means that after N=1000 there was no electrical failure. The first electri- cal failure came after 1500 cycles for both al- loys, but this was not due to a defect in the assembly (Figure 7). All defects were based on wire bond defects. After N=2000 there was a delamination with the SnAg3.5 alloy between die and solder (Fig- ure 8). That effect was temporary, which means that it wasn't for all. For example another cross feaTure aSSEmBLY maTErIaLS FOr hIGh-TEmPEraTUrE aPPLIcaTIONS continues Figure 8: Cross-section of Snag3.5 after n=2000. Figure 9: Cross-section of Snag3.5 after n=3000. Figure 7: Cross-section after n=1500 (a) hT1 and (b) Snag3.5. a b