Electrical engineer’s tiny technology has potential to revolutionize electronics

With the recruitment of luminaries like Michael Hochberg, the UW is fast becoming a nanophotonics powerhouse.

Michael Hochberg is a relatively recent transplant to Seattle, but he already blends into his new habitat: he pairs dress shirts with Adidas running shoes, and brings his dog — a large, black and enthusiastically lazy mutt named Cassandra — to the office with him.

That office is near the center of the UW’s Seattle campus, where Hochberg is an assistant professor of electrical engineering. He’s also one of the fastest-rising stars in the study of nanophotonics, the science of manipulating light at a very small scale, and his recruitment to the UW last year sets the stage for the University to become a major player in the field.

Just 28 years old, Hochberg holds B.S., M.S., and Ph.D. degrees from Caltech, and co-founded two companies — a software company, Simulant, and a hardware company, Luxtera — while still an undergraduate. His work at the UW is supported in part by an Air Force Office of Sponsored Research Young Investigators Program award.

In a sense, photonics is analogous to electronics: While electronics uses electrons to encode and transmit information, photonics relies on photons, the basic units of light. Though they might not know the term photonics, most people are familiar with the technology — fiberoptic cables, made of glass, are one example.

Hochberg works with the silicon equivalent of fiber-optic cables. They’re about 1/200th the width of a human hair. And in this field, small is powerful. “As you cram light into smaller volumes, photons interact with each other and also with materials more strongly,” Hochberg explains. What that means is that nanophotonic technology has the potential to yield computers and other devices that are smaller, lighter, faster, more powerful, cheaper and more reliable than current electronic versions.

Admittedly, all of that is still at least a few years off, says Hochberg, sitting in his office in front of a giant whiteboard that’s covered with scribbled equations. As Cassandra snoozes on the office futon, Hochberg tosses a tennis ball back and forth between his hands and explains that nanophotonics is “still in its vacuum-tube era,” referring to the primitive state of electronics before the development of the transistor-based computer chip in 1958.

In fact, creating a photonic version of a transistor — something of a Holy Grail in the field — is one aim of Hochberg’s current research. Such a device would process information perhaps 10,000 times faster than the transistors in current personal computers. Recently, Hochberg and his collaborators have taken several key steps toward an ultrafast photonic transistor by showing that light can be confined to extraordinarily small areas on silicon chips, and by building photonic circuits that work orders of magnitude faster than ever before.

Hochberg’s work also paves the way for further electronic-photonic integration, or the creation of electronic and photonic circuits on the same computer chip. According to Larry Dalton, UW professor of chemistry and electrical engineering, such integration is expected to yield “an ever-increasing array of capability” in defense, medicine (especially diagnostics) and other fields.

Hochberg speaks of wanting to create an “ecosystem” around nanophotonics in the Puget Sound region.

Dalton freely uses the word “genius” when speaking of Hochberg. It’s not just that he “can envision what other people can’t,” Dalton says, it’s that he also “has the mathematical and scientific background to test these ideas.”

For his part, Hochberg speaks of wanting to create an “ecosystem” around nanophotonics in the Puget Sound region. He’s already engaged in fruitful collaboration with Dalton’s research group, which synthesizes high-tech polymers that improve the performance of Hochberg’s silicon nanophotonic circuits. In the future, he envisions linking up with UW researchers in medicine and other fields, then broadening the circle to include the technology companies and venture capital firms of the region.

One key link in that ecosystem is an ongoing intellectual symbiosis between Hochberg and Tom Baehr-Jones, who has also joined the UW recently as a research scientist in the nanophotonics group. The two met during an introductory semiconductor fabrication course at Caltech (Tom was TA-ing the Saturday night session in which Hochberg was a student) and have been collaborating ever since, cofounding both Simulant and Luxtera and writing most of their papers together.

“Tom is the theorist, I’m the builder,” is how Hochberg explains it. The two even reflect in their physiognomy the roles that Hochberg describes: Hochberg solid, barrel-chested, and given to talking with his hands; Baehr-Jones taller, slender, with a voice that seems to float from the top of his cerebral cortex. And they sound like a pair of old college roommates — probably because they are. Encountering each other in the hallway one recent Friday, the two had an oblique, rapidfire exchange in which they agreed that the lack of dials on a piece of advanced equipment was hilariously absurd.

Another important item in the University’s growing nanophotonics toolbox, an electron beam lithography machine, will arrive at the UW next year. The “e-beam” is essentially a pencil for writing on silicon, with a point that’s just three billionths of a meter wide. It “makes building structures on the nano-scale very cheap and accessible,” Hochberg says. The instrument is being funded in part by a $1.2 million gift from the Washington Research Foundation — the largest in its history — and by the Washington Star Researchers Program. It will be the most expensive tool ever acquired by the UW College of Engineering.

The e-beam will also be the only one of its kind in the Northwest, and will enable Hochberg and others to build nano-structures right on campus. At least 30 UW faculty members have expressed interest in working with the machine, and, for a fee, outside researchers will also be able to use it. As Hochberg says, “We’re trying to lower barriers to entry to doing nano-scale work.”