Issue link: https://iconnect007.uberflip.com/i/1278896
58 PCB007 MAGAZINE I AUGUST 2020 The volume, as well as the shape of the IMC that is formed after the reflow aging, is com- parable for both processes and results in the formation of a very homogeneous and dense IMC with even grain size and without any de- fects or voids. As one of the main purposes of the ENEPIG finish is to enable a gold wire bondability with lower gold thickness, the wire bonding performance has been compared to the con- ventional pure Pd deposit. Again, the layer thicknesses were comparable, and the perfor- mance was checked throughout the entire Pd bath life. The bonding was performed using a 23 µm gold wire with a break load of 9.9 g, and the pull strength was measured in as- received condition, as well as after aging for four hours at 150°C. The mean pull values that were measured were comparable for both processes and consistent throughout the bath life (Figure 6). Only fracture modes of 2, 3, and 4 were ob- served; no indication for lift-offs or peeling could be detected. The tests were performed over a Pd bath life of Pd-Core® of 16 MTO @ 0.5 g/l Pd, which equals eight MTO of the con- ventional pure Pd bath with a Pd content of 1 g/l. Therefore, it can be concluded that the performance of both processes is fully com- parable and that the Pd-Core® offers the same reliability as the mass production proven con- ventional pure Pd deposit. Process Robustness Earlier studies showed that the Pd plating process could be quite sensitive to any pro- cess contamination or poor rinsing [1–2] . The rinse quality of the rinse step between the nickel and the palladium plating can have a significant impact on the initial seeding of the palladium layer, which can influence the deposition rate as well as the initial crystal structure and layer growth of the deposit. To check the sensitivity of the palladium plat- ing towards the rinse quality, an additional test was performed where the rinse between nickel and palladium was contaminated on purpose. The rinse water was doped with nickel bath solution up to a conductivi- ty of 500 µS. The test panels were pushed through the whole process until the Pd step and rinsed and dried right after the palladi- um plating. The palladium deposit was in- vestigated by means of surface SEM on eight different areas on the test coupons. Pores in the Pd layer top view were counted, and the Figure 6: Pull strengths and fracture modes for gold wire bonding of ENEPIG.