Skip to Content

Adding Speed to Silicon

A new way to grow various semiconductors on silicon could speed up electronics.
July 24, 2007

Silicon is the material of choice in the electronics industry, but there are other semiconductors that, if grown on a silicon wafer, could make transistors twice as fast. The problem is, these semiconductors are largely incompatible with silicon. But now a startup called AmberWave Systems has developed a new method for matching the different materials that could lead to faster, smaller electronics. Additionally, combining the different types of semiconductors could lead to cheaper lasers and other photonic devices, since the semiconductors used to make lasers and detectors are more expensive.

Fixing fissures: Usually when semiconductors, such as germanium, are grown on silicon, they crack due to strain caused by a crystal mismatch between the two materials. AmberWave is proposing a new way to grow germanium and other semiconductors on silicon so that they don’t crack. Above, germanium is grown on an eight-inch silicon wafer.

For decades, researchers have been trying to effectively grow various types of nonsilicon semiconductors on silicon, but so far, no technique has made it much farther than the lab. The fundamental problem, says Anthony Lochtefeld, AmberWave’s vice president of research, is that atoms in silicon are spaced closer together than are those in semiconductors such as germanium, gallium arsenide, and indium phosphide. “The [atomic] lattices don’t fit,” says Lochtefeld, and the materials physically crack, which “prevents you from making anything useful.”

In research presented at the SEMICON West conference in San Francisco last week, Lochtefeld showed how his team carved 500-nanometer-deep trenches (they range from about 250 to 400 nanometers wide) into a layer of silicon dioxide on silicon and then filled the trenches with germanium. At the bottom of the trench, the germanium predictably fissured where it contacted the silicon. However, due to the innate orientation of germanium’s crystal lattice, as well as that of silicon, these cracks only crept up to about half the height of the silicon dioxide trench. “With a high enough sidewall,” says Lochtefeld, “you can trap the dislocations.” Above the trench, usable germanium was able to grow relatively defect-free.

AmberWave’s work is an “exciting development,” says John Bowers, a professor of electrical and computer engineering at the University of California, Santa Barbara. “Their approach is promising and will hopefully yield material that’s useful.”

However, at this point, AmberWave’s technology is still in its early stages, and the researchers have yet to prove that their method will yield working devices. Right now, the team is focused on testing other semiconductor materials, such as indium phosphide, which is used to make infrared lasers found in telecommunications networks. Lochtefeld doesn’t expect to encounter surprises even though its lattice has larger spacing than that of germanium and gallium arsenide, another material successfully tested, due to positive preliminary results. But before the technology can be used to build transistors or photonic devices, there is still more work to do to ensure that the defects for each material are minimized.

Lochtefeld is optimistic that AmberWave’s technology will find its way into transistors and photonics as early as 2012. The dimensions of transistors are shrinking, and silicon, as it’s used today to make these transistors, will not be able to scale down and maintain the same speed. The industry estimates that silicon will reach its physical limitations by about 2012, and there’s a scramble to find the best materials to replace it. (See “Beyond Silicon.”) “People have been going after this for 20 years,” says Lochtefeld, “and for some very important applications, we think we’ve found a way to make this work.”

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.

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.

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.

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.