Customer issue: The customer wanted to obtain wire bond pull data on immersed silver substrates. The substrates were to be baked at varying lengths of time (0, 2, 4 and 8 hours at 125 degrees Celsius) prior to processing and testing.
Investigation Technique
Each substrate was labeled on the back side as to the length of time for baking prior to performing the wire bonding.

Wire Bonding

• The boards to be tested were 6.1 cm [2 7/16 in.] by 3.8 cm [1.5 in.].
• Wire bonds were made using a FEK Delvotec 5430 at 102kHz, and a bond weight of 15.
• 200 bonds using 1.25 mil diameter aluminum wire with a loop height of 3.50 mil were tested.

Pull Testing
Wire bond pull tests were performed on a Dage model 4000 at K&S in Horsham, PA using the following parameters:

• 100g cartridge (2013480)
• Test speed of 700m/sec
• Test load of 20g
• Fall back of 30%

Results
A summary of the wire bond pull testing results is displayed in Table 3-1.
Wire Bond Fracture Locations
Each test board possessed four wire bonds for a total of 8 contact bonds per board. All 8 of these locations were examined under a 400X optical microscope on three boards from each aging time. In addition, a scanning electron microscope (SEM) was used to examine the fractured areas on two of the boards.

• After wire bond pull testing, the fractures were noted to occur within the wire at all locations except for about 40% of the wire bond source sites where the fractures occurred at the wire/pad interface.
• When pull testing a strong wire bond, the break should occur in the wire, leaving behind a remnant from the wire-to-pad bond (Figure 5-1). This situation was observed to occur on 47 out of the 48 destination bond sites which were inspected from various aging times. Inspection of the wire bond source sites (Figure 5-2) revealed that 20 out of 48 wire bond sites inspected did not show any wire remnant, indicating that the separation occurred at the wire/pad interface. Table 3-2 lists the inspection results per aging time. Listed are the number of bond sites that visibly exhibited failure at the pad surface.

Conclusions
The fracture mechanism was not influenced by the aging time, but was affected by bond location. The destination sites exhibited failure within the wire in nearly every instance. One possible reason for this could be that the wire near the destination site was crimped and positioned at a greater angle than the wire at the source site. This created a stress concentration point that could have made the wire more susceptible to fracture.

Recommendations
To ensure the bondability and reliability of wire bonds, the bonding surface must be free of any contaminants. Cleaning is the fundamental process needed prior to wire bonding. Optimum bonding parameters are the key for reliability and maintaining high yields.

The key variables that affect bonding include:

• Bonding force and pressure uniformity
• Specified bonding temperature
• Bonding time
• Ultrasonic frequency and power