The Nobel Prize and Its Discontents
Some imaging pioneers have a problem with the Nobel committee
Photo-Illustration: Sean McCabe; Photos, from left: Eugene I. Gordon, Alcatel-Lucent/Bell Labs (3); CCD chip: Michael F. Tompsett
Bell's Best: Eugene I. Gordon, Michael F. Tompsett, Willard S. Boyle, and George E. Smith [from left] each had a hand in the CCD.
The Nobel Prize is the peak honor in physics. Yet this year it celebrated not science, but technology: "the invention of an imaging semiconductor circuit—the CCD sensor." The winners were two IEEE Fellows from Bell Telephone Laboratories, Willard S. Boyle and George E. Smith.
Two other IEEE Fellows and former Bell Labs colleagues, Michael F. Tompsett and Eugene I. Gordon, say the Nobel committee has made a mistake. Their complaints reveal a lot about how inventions happen and how credit for them is given. At the heart of their complaints is that the Nobel Prize winners had little to do with the charge-coupled device's use in imaging—the reason it became so important to astronomy and consumer electronics.
According to Smith, in 1969, he and Boyle were seeking a memory technology. Magnetic bubble memory had recently been developed at Bell Labs, and it was all the rage. That technology stored data as a linear pattern of tiny magnetized spots, called bubbles, on a ribbon of magnetic material. Current passed through the ribbon would push the bubbles along in one direction so they could be read out or written sequentially.
Boyle, then executive director for the Bell Labs area that worked on silicon, worried that support would be diverted from silicon research toward magnetic bubble memory. He asked his friend George Smith to help him come up with a competitor.
On a chalkboard they sketched out a concept similar to bubble memory that used charge instead of magnetic domains. The idea was to build a line of potential wells in silicon by creating capacitors out of silicon, silicon oxide, and metal electrodes. Charge—directly injected in the early CCDs—would accumulate in the wells. By applying alternating voltages, you could then move the accumulated charges from one well to the next until they reached the edge of the chip, where the amount of charge could be read out and digitized.
Both Boyle and Smith say that the device's potential for imaging was obvious to them at the time, and Smith has notebook entries to prove it. But they did not pursue that application. The CCD continued its journey at Bell Labs, under the guidance of Michael Tompsett.
Tompsett's problem with Boyle and Smith's Nobel award comes down to the citation, which includes the word imaging. And the CCD imager was invented by Tompsett, not by Boyle and Smith. So says not just Tompsett but also the United States Patent and Trademark Office.
Tompsett, who ran Bell Labs' CCD group in the 1970s, is the sole inventor listed on U.S. Patent No. 4085456, "charge transfer imaging devices." The patent, filed in 1971, covers linear scanners and area imagers.
Tompsett's key invention was a scheme called frame transfer. The invention solved a big problem with using a CCD as an imager: The CCD continued to sense light and gather charge even as each line of pixels was read out, smearing the image in the direction of the charge transfer. Tompsett's idea was to duplicate the entire CCD structure on a part of the chip that wasn't exposed to the image. He found a way to rapidly transfer the charge collected in the imaging CCD to the hidden CCD. The image was then read out from the hidden CCD, while the imaging side took another picture.
Tompsett has no illusions about what will happen in Stockholm this month. "You're not going to change [who wins] the Nobel," he says, but "it would be nice to at least share the credit." The citation for the award, he says, should be altered to eliminate any mention of imaging.
Smith disagrees. Tompsett "can be credited with good engineering but not with the concept of imaging with a CCD," he says.
Gordon led the development of the CCD's precursor technology, the silicon-diode-array camera tube, which Bell Labs developed for an ultimately abandoned product, the Picturephone. In that device, light fell on the silicon-diode array, producing pockets of charge. Then a low-energy electron beam scanned the array, kicking out current in proportion to the intensity of the light.
With the Picturephone camera already in production, Gordon was not interested in a new imager. But Tompsett pushed. "If it were not for Mike's perseverance, Bell Labs would not have done any imaging work with the CCD," says Gordon.
So who deserves the accolades? "It depends on what you're celebrating," says Carlo Séquin, who joined the Bell Labs CCD team about nine months after the project began.
"My initial assumption was the Nobel in physics goes to fundamental concepts," says Séquin, now a professor of electrical engineering and computer science at the University of California, Berkeley. "If the fundamental concept was the charge-transfer principle, then that goes to Boyle and Smith, and maybe Gene Gordon." If it's the invention of a practical CCD imager, "credit would go to Mike Tompsett, and possibly Gilbert Amelio," he says. (Amelio led commercial CCD development at Fairchild Semiconductor.)
It's easy to see why the Nobel committee went with Boyle and Smith. The CCD is synonymous with its only practical application: imaging. And according to many authoritative sources, Boyle and Smith invented the CCD. But had the Nobel nominators looked one step down the chain of invention, things might have been different.