Here is the abstract you requested from the imaps_2018 technical program page. This is the original abstract submitted by the author. Any changes to the technical content of the final manuscript published by IMAPS or the presentation that is given during the event is done by the author, not IMAPS.
|Laserbonding instead of ultrasonic wire bonding – an alternative joining technology for power applications|
|Keywords: Power Electronics, Laser Processes, Interconnects|
|Ultrasonic wirebonding is reaching its limits of processable wire and ribbon sizes for the high currents which are required in battery modules for e-vehicles. While bonding many wires in parallel is technically feasible, it is not attractive economically. Laser welding as a corresponding technology can process connectors of practically unlimited cross-section and hence current-carrying capability, but it is far less flexible than wirebonding because it requires pre-fabricated connectors which have to be placed in a separate step, and there are no simple means of correcting for positional inaccuracies of the battery cells. A technology combining the advantages of wirebonding and laser welding would be desirable and is described here. A fibre laser head (500 to 1000 W power) is mounted on a standard heavy-wire bonder and extends the ultrasonic system. Because the infrared radiation of the laser (approx. 1060 nm) is poorly absorbed by copper surfaces at room temperature and therefore the welding process would be non-uniform and highly dependent on the quality or oxidation state of the coper surfaces, it is crucial to employ highly brilliant laser sources which can be focused to a small focal spot of about 30 µm diameter. At this high intensity, deep penetration welding is easily accomplished with excellent control of the weld seam. Due to the use of a galvanometer scanner in the laser optics, oscillation welding is possible which allows to shape the weld seam to any desired width, while keeping the weld penetration extremely shallow and independent of the depth. This results in excellent control over the connection cross section and makes it possible to weld relatively thick ribbons onto a much thinner substrate. Because the laser beam moves in a circular fashion while advancing more slowly, a large share of the heat is dissipated inward of the circling pattern and so melts the area far more efficiently than a conventional laser weld. This permits creating a large interconnection area with far lower heat input and hence thermal stress than conventional laser welding. Apart from the excellent thermal properties, laser bonding enjoys two further advantages: first, the bonding surfaces do not need to be as clean as it is required for classical ultrasonic bonding; secondly, the clamping is far simpler. The surfaces of both ribbon and bonded area can be of lower quality than required for ultrasonic bonding because there is no friction required to form the bond. Laser bonding proceeds through a molten phase just like any laser welding and hence can tolerate a fair amount of surface contamination, especially organic contaminants which are decomposed during the process. Because there is no controlled friction motion under pressure, there is no need for a rigid clamping system like in ultrasonic bonding. The laser bonder does, however, operate with a touchdown system to detect the exact height position of the bond surface and exerts a programmable bond force, just like in ultrasonic bonding. This makes sure that there is no welding gap between ribbon and bond pad and again ensures excellent uniformity of the weld at low energy input. Welding seams have been achieved on standard 18650 battery cells without any thermal damage to the internal components due to the excellent depth control of the laser beam. The shear strength of the seam increases with the oscillation amplitude, i.e. with increasing width of the weld seam, while holding the total imparted laser energy constant. This is particularly valuable as copper ribbons of increasing width and thickness are desired to be connected to larger battery cells to allow currents beyond 100 A. Current bonder versions are already capable of welding copper ribbons of 10 mm width and 0.5 mm thickness which have a fuse current of well above 400 A. A highly attractive application concerns secondary connections from power modules, i.e. the terminals leading from the DCB (Direct Copper Bonding) substrate to the outside of the package. Today these terminals are usually soldered to the DCB plate, or they are mounted to the package frame and connected to the DCB by multiple wire bonds. A newer connection method uses high-power ultrasonic welding of terminals to the DCB. All of these methods are fairly costly and technologically difficult, especially the ultrasonic bonding which is a serial process and runs the risk of cracks in the substrate. Laser welding the terminals directly onto the DCB copper layer is, in essence, done by running the Laser bonder without a ribbon, but just as a laser welder. The difference is that the bonder can press the terminals against the DCB surface with a programmed force and it can monitor the height at which the terminal is mounted, so as to make sure the weld is executed correctly. Standard laser welding is not capable of controlling this. A further development step which exploits this excellent depth control will cover the bonding of semiconductor chips by laser bonding. Today, bonding high-power semiconductors is moving towards copper wire but a number of process problems are slowing the packaging development. We believe that some variants of laserbonding will be capable of achieving this at lower cost. A joint German Research Project is already under way to address this technology.|
F&K DELVOTEC GmbH