Thermosonic wire-bonding is the most common technology used for creating first-level interconnections in microelectronics. The method combines pressure, elevated temperature and ultrasonic energy for forming bonds. Based on the literature survey, only a few research works were carried out about the effect of firing time (Electric Flame Off time) on the size and mechanical properties of the formed ball (FAB – Free Air Ball). My research focuses on thermosonic wire-bonding using gold wires, but the technology can use copper wires as well. In fine-pitch electronics, the components are getting smaller and smaller, and the bonding technology needs to keep up with the issues arisen from the decreasing components sizes.
The electrical, mechanical and size properties of ball (first) bonds depend on multiple factors, and the firing time is one of the major ones. There are many criteria, which the optimal process parameters depend on, like the diameter of the wire, and the mechanical property of the ball bond. The last is tested by pull- and shear tests.
I performed literature survey on the topic of thermosonic wire bonding technology, on its role in wire-bonding in microelectronics, and on how other similar first-level technologies work. The main topic of my thesis work was the thermosonic wire bonding, so I also studied the operating principle of plasma cleaning belonging to this technology. Applying plasma cleaning is absolutely necessary for increasing the reliability of the interconnections. I also researched the critical parameters which are significant in making a good interconnection between the semiconductors and the substrate pad.
My task was to test gold wires with different diameters with multiple wire bonding machines, and to determine what firing time parameters are required to produce the optimal free air ball diameters. I measured the free air balls with optical microscope, and performed shear tests to characterise the mechanical propertied of the ball bonds.
I used three type of gold wire, which had the diameters of: 25, 32 and 50 µm. The 25, 32 and 50 µm wires were tested between 205–605, 435–835 and 750–1350 µs firing times. There were 12, 12 and 11 examination points, to see how the free air ball diameter changes depending on the firing time. I made 40 samples for every firing time, and I calculated the average of the values. For the shear test, I made 120 ball bonds for the chosen firing times. For the 25 µm, and the 50 µm wires I used 7 and 7, for the 32 µm 6 measurement points.
The results showed that there is a linear correlation between the firing time, and the diameter of the free air ball. For the 25 µm gold wire, 295–315 µs was the firing time window, which were given good results, for the 32 µm wire, it was 510–585 µs, and for the 50 µm gold wire, the good result were between 950, and 1075 µs.