Issue link: https://iconnect007.uberflip.com/i/770652
20 The PCB Magazine • January 2017 ENEPIG deposits of varying gold thicknesses from various suppliers were wire bond tested. It is believed the samples for the IPC study were not baked to simulate die attach and this current study attempted to determine if any effects of baking at 175°C for 16 hours were seen on gold wire bond results when using different types of immersion gold. A depiction of the typical wire break modes is shown in Figure 6. Ideally, wire breaks at locations B, C, and D are preferred. Wire breaks at position A and E would indicate a poor bond between wire and substrate. Results for the current study are shown in Table 4 and Figure 7, respectively. Despite heat aging, all results appeared acceptable, well above the 3-gram pull strength minimum. The standard displacement immersion gold sample with thick deposit showed lower minimum wire pull strength and higher variation in values when compared to the other samples. Wire breaks in all tests were in the wire either just above the first bond or just above the second bond (modes B or D). No failures were seen at the wire/substrate interface (modes A or E). SEM Examination of the Deposits Figure 8 shows the ENEPIG deposit top- down. In general, the palladium and gold uni- formly coat the electroless nickel. Experience has shown that attempts at ob- taining very thick gold with ENEPIG using a standard displacement immersion gold results in some damage in the electroless nickel layer below the electroless palladium deposit. Of the three deposited metals the electroless nickel is the least noble. Any access the immersion gold may find to the nickel will result in nickel dis- solution into the immersion gold solution. Excess dwell time in the immersion gold solu- STUDY OF IMMERSION GOLD PROCESSES USED FOR BOTH ENIG AND ENEPIG Figure 5: Wetting balance test results, sample 4 after three reflows. Table 4: Wire bond test result data for the six samples. Figure 6: Wire break modes. Figure 7: Wire bond data, 20 points per sample.