.........that connect integrated chips through a direct-write approach. The innovative technique is similar to writing with a fountain pen in three-dimensional space, and could have the potential to shrink integrated circuits and expand microelectronics.
The team are using a micropipette with a copper electrolyte solution – when the pipette comes into close contact with the surface, a liquid bridge forms between the end of the pipette and the bonding pad. An electric current is then applied, causing the copper in the solution to deposit as solid metal. As the end of the pipette tip moves through space, copper will continue to deposit from the solution in the pipette, in the same way as ink from a pen, thus creating a wire.
The research, published in Science [Hu, J., Yu, M., Science (2010) DOI: 10.1126/science.1190496], has demonstrated up to 20 of the new wires bonded to a single standard bonding site. The study had to calculate the best speed to move the pipette tip to maintain the liquid bridge between the nozzle and the growing wire, and to understand how to “write” the wires laterally for chip-to-chip bonding.
Another challenge was that the bonding pad for traditional wire bonds takes up a quite a bit of room, which makes the act of shrinking wiring difficult. With electronic devices getting smaller all the time, it is the size of the required 50-by-50 micron square bonding site that is holding back innovation, and integrated functions on a very small scale.
With no existing cost-effective technology that allows you to wire-bond microstructures, the team has got rid of those wires, and instead directly produced them on-site between the connection points. Min-Feng Yu, a professor of mechanical science and engineering at the University of Illinois, who lead the study, said “This technique means the pads can be much smaller than what’s needed for traditional wire-bonding technology.” This size reduction could mean that manufacturers are able to produce more chips per wafer of semiconductor material, and could allow more complex integrated functions in microelectronics.
The team now hope to develop a scaled-up system that can perform batch fabrication of such 3-D microscale/nanoscale metal structures, and perhaps explore further shrinking of structures down to sub-100 nm dimensions.
Having demonstrated their technique using both copper and platinum wires, they now plan try the technique with other metals, and with the process now automated, they are also looking to develop arrays of micropipettes that produce wire bonds in large quantities for more efficient manufacturing.