Computing power has increased exponentially over the last few decades due in part to advances in optical lithography, which allows manufacturers to create patterns on computer chips that guide the placement of smaller and smaller transistors.
As transistors shrink closer to the atomic scale, however, optical lithography is reaching its natural limit. Over the last 20 years, researchers have experimented with new methods of patterning that use self-assembling block copolymers, which can create structures in the sub-10-nanometer scale. Now, the discovery of a new strategy for assembling block copolymers by University of Wisconsin-Madison materials engineers is bringing that technique closer to wide-scale use.
“Boundary-directed epitaxy of block polymers is an alternative technique to make these nano-scale patterns with very fine features, and it has the potential to be cost effective,” says Robert Jacobberger, who conducted research in the Department of Materials Science and Engineering until August of 2020.
Previous strategies for creating nanoscale patterns using block copolymers—graphoepitaxy and chemoepitaxy—require researchers to fabricate trenches or ultra-narrow guide stripes that cause the polymers to assemble in certain regions. Creating and removing those structures, however, leads to several extra, often technically difficult steps.
Boundary-directed epitaxy simplifies the process and results in alternating stripes of vertical polymers which are useful in creating the nanopatterns. In this process, templates consisting only of flat, relatively large features are used to produce more complex block copolymer patterns with enhanced feature resolution, circumventing the need for trenches and ultra-narrow guide stripes.
To realize boundary-directed epitaxy, the team invented a new type of template comprised of relatively wide 80-nanometer ribbons of graphene, a one-atom-thick layer of carbon atoms, on top of germanium, which presents minimal topography due to graphene’s atomic thinness and results in abrupt transitions in surface composition at graphene’s atomically sharp edges. They then coated the wafer with an ABA triblock copolymer in which the A polymer is poly(propylene carbonate) and B is polystyrene. The A block prefers to assemble on germanium and the B block prefers the graphene.
As they heated the wafer, they discovered that vertically oriented, narrow stripes of polymer became pinned at the boundaries between the graphene ribbon and the germanium substrate. Additional stripes of polymer then pushed in toward the center of the graphene ribbon, where each successive stripe was guided by the first stripes locked at the boundaries. After a short time, orderly, alternating parallel lines of 6.4-nanometer-wide polymers covered the graphene.
In this manner, a single, planar, relatively-wide 80-nanometer graphene ribbon was used to align, position and order significantly finer 6.4-nanometer block copolymer features that are beyond the resolution of conventional lithography.
The team also showed it was possible to use the technique to produce bends, jogs, T-junctions and other shapes important in creating fully functional patterns for electronic devices.
The low temperatures required by the technique, just 160 degrees Celsius, and short assembly period of 10 minutes, make it compatible with current fabrication techniques, though Jacobberger estimates that it will be at least five or 10 years before the technique is ready for industry use, since many obstacles still need to be overcome for its integration.
While graphene and germanium are not commonly used in the semiconductor industry, Jacobberger says boundary-directed epitaxy works with other materials as well, making it more compatible with mass production.
“Demonstrating this process with other template materials has almost endless possibilities,” he says. “There’s also a long list of block copolymers that could be used that offer advantages such as even smaller feature sizes.”
The assembly technology is currently being patented through the Wisconsin Alumni Research Foundation.
Jacobberger and Arnold worked closely with their collaborators, including Shisheng Xiong at Fudan University, Vikram Thapar and Su-Mi Hur at Chonnam National University, Paul Nealey at the University of Chicago, Guang-Peng Wu at Zhejiang University, and Tzu-Hsuan Chang at National Taiwan University in addition to Zhenqiang Ma, Vivek Saraswat, Austin Way and Katherine Jinkins at UW-Madison, each of whom contributed important findings to the study.
The research was supported by the Department of Energy’s Basic Energy Sciences program and the National Science Foundation, among other agencies.