Issue link: https://iconnect007.uberflip.com/i/1533339
MARCH 2025 I PCB007 MAGAZINE 77 Figure 10: An example of the Ni-less surface finish intermetallics after six reflow cycles and 1,000 hours of aging. Table 2: Comparison between previously published data on DIG intermetallics 10,11 and the Ni-less surface finish intermetallics. DIG surface finishes have no barrier between the copper and gold to prevent interdiffusion and, therefore, intermetallic growth. Since the DIG intermetallics are thicker than the Ni-less surface finish intermetallics, the solder balls that are on the Ni-less surface finish should be stronger than those on DIG surface finishes. Table 3: Comparison between previously published data on EPIG intermetallics 12,13 and the Ni-less surface finish. tently thinner intermetallics than the EPIG/ EPAG surface finishes. is shows that the bar- rier layer in the Ni-less surface finish performs better than 100 nm of palladium in inhibiting intermetallics growth. Since the EPIG/EPAG intermetallics are thicker than the Ni-less sur- face finish intermetallics, the solder balls that are on the Ni-less surface finish should be stron- ger than those on EPIG/EPAG surface finish. In conclusion, the Ni-less surface finish with the nano-engineered barrier layer inhib- ited intermetallic growth compared with sur- face finishes with no barrier between gold and copper (DIG) and with 100 nm palladium bar- rier between gold and copper (EPIG/EPAG). e thinner intermetallic areas in solder balls on the Ni-less surface finish should result in strong solder joints, which was tested in the next section. Solder Joint Strength: Solder Ball Pull and Shear Test e figure below shows the different types of failure modes during pull tests and shear tests on solder balls on the Ni-less surface finish. e Intermetallics: EPIG/EPAG Comparison is comparison was conducted to see how the barrier layer component of the Ni-less sur- face finish compares to a barrier layer in a dif- ferent surface finish, in this case 100 nm of pal- ladium in EPIG/EPAG surface finishes. e references for the EPIG/EPAG sur- face finishes in Table 3 differ slightly in sample preparation, with one of the references only using 300 hours of heat storage instead of 500 hours like the Ni-less surface finish samples and the other EPAG reference. Additionally, both of the EPIG/EPAG references only sub- jected the samples to one reflow cycle, com- pared with the six reflow cycles of the Ni-less surface finish samples. Both EPIG/EPAG ref- erences used 100 nm of palladium and 100 nm of gold. Despite the harsher sample preparation for the Ni-less surface finish samples, the results show that the Ni-less surface finish had consis-