Skip to Content

3-D Transistors Made with Molecular Self-Assembly

Researchers at IBM have made the first 3-D transistors using a promising new manufacturing approach.

A new way of building computer chips is taking shape that involves synthesizing molecules so that they automatically assemble into complex structures—which then serve as templates for etching nanoscale circuitry into silicon. The approach could let the computer industry continue to shrink electronics beyond the resolution of existing manufacturing machinery. IBM researchers have been the first to make speedy 3-D transistors using this new method.

In 2011, the computer industry adopted 3-D transistors for high-end integrated circuits because they switch faster while consuming less power than planar ones (see “How Three Dimensional Transistors Went from Lab to Fab”). Such circuits have conventionally been made via photolithography, the same process used for most computer circuits. In this process, silicon wafers are coated with a light-sensitive material called a photoresist, and then are exposed to a pattern made by shining light through a filter known as a mask. Wherever light strikes, the photoresist cures; the rest is washed away, and the wafer is then etched chemically to create features in exposed parts of the surface.

For the fastest microchips, which have elements as small as 22 nanometers and 80 nanometers apart, this process is repeated about 30 times, says Kwok Ng, director of nanomanufacturing at the Semiconductor Research Corp. in North Carolina. Each step requires its own expensive mask, and every step adds time to the process. Photolithography will work for the next generation of chips, with features 14 nanometers in size. But for faster chips with smaller features, photolithography will become too expensive and complicated, and will run into limits determined by the wavelength of light.

The IBM group used a new approach known as “directed self-assembly,” using a class of materials called block copolymers (polymer chains are made up of two kinds of monomers, or blocks).

It is possible to make these materials self-assemble into complex patterns, such as a densely packed row for stripes. This is done by tailoring the polymers’ length, size, and other characteristics, such as how two blocks attract and repel one another.

Patterns made in this way can be much denser than what is possible using lithography. That means the approach can be used to create the smallest, most densely packed, and uniform parts of an integrated circuit: for example, the channels of silicon transistors, or the fins in 3-D transistors. The rest of the circuit would still be formed using the conventional methods.

The IBM group used existing photolithography methods to prepattern a photoresist coating to form a series of deep, parallel trenches. These trenches then help direct the assembly of block copolymers, which are arranged in patterns needed to etch transistor fins that were smaller and more densely packed together than is possible with photolithography alone. The resulting working devices had features as close together as 29 nanometers, far smaller than the 80 nanometers that is currently possible.

“They’ve used these polymers not just to make pretty patterns but to make some working devices,” says Caroline Ross, a materials scientist at MIT who is researching directed self-assembly. “They’ve demonstrated a creative way of getting patterns that normally wouldn’t form.”

Directed self-assembly is already being tested by some chipmakers, says Ng. However, block copolymers tend to assemble with some defects, and it remains to be seen whether the process can be controlled well enough at large volume. 

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.