Building nanowires to connect future chips

As computer chips are getting smaller and smaller, it becomes crucial to find new ways to connect them. Canadian researchers have developed a method to produce self-assembled nanowires that are 5,000 times longer than they are wide. They've been able to obtain nanowires measuring only 10 nanometers in width, but with a length of 50 microns. These nanowires could be used in about 10 years or more when the semiconductor industry reaches the 10-nanometer limit.

As computer chips are getting smaller and smaller, it becomes crucial to find new ways to connect them. Canadian researchers have developed a method to produce self-assembled nanowires that are 5,000 times longer than they are wide. They've been able to obtain nanowires measuring only 10 nanometers in width, but with a length of 50 microns. These nanowires could be used in about 10 years or more when the semiconductor industry reaches the 10-nanometer limit.

Self-assembly of nanowires

The figure above summarizes the chemistry used to produce these metallic nanowires (Credit: University of Alberta).

This research project has been led by Jillian Buriak, Professor of Chemistry at the University of Alberta and senior research officer at the National Institute for Nanotechnology (NINT). She was helped by her colleagues of her research group, including Dr. Dong Wang, a NSERC Visiting Fellow.

Here is some details about the process given by the University of Alberta. "In one example, 25 parallel platinum nano-wires were made using this self assembly process, with each wire measuring only 10 nm in width, but extending to a length of 50 microns. While the idea of wires ‘self-assembling’ sounds like something from science-fiction, it’s a natural process, says Buriak. “You are the product of self-assembly. The way DNA forms a double helix is self-assembly. It’s just that molecules will recognize each other, bind to each other and then they’ll form structures,” she said. “And the molecules we’re using are actually very simple. They’re just polymers, just plastics that do that naturally.”"

In "Self-assembly extends to 10 nanometers," EE Times delivers more technical details about the process. "Conventional lithographic tools first fabricated a micron-sized data bus, followed by the nanoscale patterning of individual 10-nm-wide wires, which were inserted into the bus using self-assembling block copolymers. Different block copolymers can self-assemble any repeating pattern--from nanowires for interconnections to nanoparticles for ultra-dense flash memories."

R. Colin Johnson, the author of the article, asked Buriak why she was using copolymers. "We have been using self-assembling block copolymers for a while -- mostly to make nanoparticles for floating gates -- but this is the first time we have been able to use them as templates for continuous nanoscale wires," she said. And she added, "The reason we use block copolymers is because they are compatible with current silicon manufacturing techniques. We made as few changes as possible from traditional chip processing -- you just use a specific copolymer as if it were a polymer, and they automatically self-assemble into the kind of nanoscale structures that you want -- like parallel lines for buses, or nanoparticles for floating gates."

Now, let's return to the University of Alberta news release to discover why these findings could mean for the future. "While the new process could provide the solution for computer manufacturers looking for ways of increasing the speed and storage capacity of electronics, it could also mean cheaper electronics as well. “If you have to go and lithographically define one single wire, it’s going to be painstakingly hard and expensive,” said Buriak. “But, if you can have a cheap molecule do it for you, that’s great, that’s going to be much cheaper, use much less energy and be a little more environmentally friendly.”"

For more information, this research work has been published in Nature Nanotechnology under the name "Assembly of aligned linear metallic patterns on silicon" (August 2007, Volume 2, Number 8, Pages 500-506). Here is the beginning of the abstract. "In order to harness the potential of block copolymers to produce nanoscale structures that can be integrated with existing silicon-based technologies, there is a need for compatible chemistries. Block copolymer nanostructures can form a wide variety of two-dimensional patterns, and can be controlled to present long-range order. Here we use the acid-responsive nature of self-assembled monolayers of aligned, horizontal block copolymer cylinders for metal loading with simple aqueous solutions of anionic metal complexes, followed by brief plasma treatment to simultaneously remove the block copolymer and produce metallic nanostructures."

The full text of this article is available as an HTML file or in PDF format (7 pages, 912 KB). The illustration above has been extracted from this paper.

Sources: University of Alberta news release, via EurekAlert!, August 28, 2007; R. Colin Johnson, EE Times, August 7, 2007; and various websites

You'll find related stories by following the links below.