Today in more than 90 % of the cases the integrated circuits and other semiconductors are connected with the substrate by wire bonding technology. Wire bonding is commonly used because of its flexibility and high reliability. During thermosonic wire bonding pressure, ultrasonic energy and thermal energy are used to make the bonding joints. The most commonly used wire material is gold because it has high reliability and high oxidation resistance, although the price of gold in the last decade became very expensive.
There are several parameters that have an influence on the quality of the wire bond and it is important to optimise these parameters. One of these parameter is the used temperature on the substrate. The objective of my studies was to examine the effect of pre-heating’s number on the bonded joints in two material structures.
First I researched the most important parameters and procedure of thermosonic wire bonding. I also made a researched about the quality test of wire bonding, the gold-aluminum intermetallic layer and its reliability and the copper wire bonding and its future uses. After these researches I made destructive pull and shear tests on the samples. The samples were pre-heated different times so I could make a connection between the number of pre-heats and the measured shear strengths.
Before the destructive tests the specimens were aged in a thermal shock chamber for 1000 cycle. During pull test I didn't find any relation between the number of heating and the quality of the joints, although the wires weakened from the ageing process. The failures of the non-aged wires were at the neck of the ball bond and at the heel of the wedge. After 1000 cycles the failures were mostly due to the break of the loop. The results of the shear test showed that the heating made the bonds stronger in both gold-gold and gold-aluminum structures. Expected results were observed during the aging process from both material structures that have been researched previously.
The gold-aluminum intermetallic layer first was inspected by cross-section. Neither the optical microscopy nor scanning electron microscopy found any evidence of the intermetallic layer because it was too thin. Therefore I dissolved the aluminum pad under the gold bond with KOH solution. With this method I was able to observe the coverage, the thickness and the roughness of the intermetallic layer.