IBM Is Using Tiny Tubes to Grow the Chips of the Future

Carbon nanotube are too tiny to carve. So IBM is using chemistry to grow chips like crystals.
NanotubeTA1.jpg
IBM Research

The reign of silicon may be coming to an end. For years, researchers and entrepreneurs hoped that carbon nanotubes would revolutionize microchip design. These tiny, molecular-level structures could, in theory, be used to make chips that are six to ten times faster than today's silicon-based variety—and use far less electricity.

In addition to faster, more efficient chips for laptops and smartphones, tiny but powerful processors could enable new types of technology, such as bendable computers and injectable microchips, or nano-machines that could target cancers in the body.

Now a team of IBM researchers say they've made a breakthrough that brings the nano-dreams of the past closer to reality.

The problem with nanotubes has always been their size. By definition, nanomaterials are incredibly small, which makes them incredibly hard to work with. Traditionally chip makers take a piece of silicon and essentially carve it into what they want--not unlike the way sculptors take slabs of stone and chisel them into the shapes they want. Chiseling patterns into carbon nanotubes, however, has proven incredibly difficult.

IBM Research

"The analogy I use is that it's like building a statue out of a pile of blocks," says IBM Research materials scientist George Tulevski, who will explain the new process tomorrow at TED@IBM. "You can't place these nanotubes one by one into the pattern you want."

Instead, Tulevski and his team have figured out a way to "coax" the nanotubes into specific structures using chemistry. Instead of a "top-down" approach of trying to place each block manually, the team is treating the nanotubes with chemicals that cause them to assemble themselves into new structures. It's a bit more like growing a crystal than carving a statue.

"We're trying to tackle that problem by borrowing from nature, because nature builds everything that way," Tulevski says. "We think that's one of the key missing pieces."

Nanotechnology, a darling of 1980s and `90s science fiction, has been making something of a comeback in recent years. Tulevski's breakthrough follows another milestone passed by IBM Research last year when a team developed a new way to pack more carbon nanotube transistors into a smaller space. Meanwhile, companies like Nanotronics Imaging have developed microscopes that could make it easier to manufacture nano-scale devices.

But the IBM Research team is still years away from being able to manufacture nanotube-based chips at scale. And because silicon chips are still getting faster, the IBM team needs to not only create a process for reliably manufacturing nanotube-based processors, but to make the processors faster than silicon chips will be in a decade. Let's just hope it doesn't take so long this time to get there.