Who invented the CCD for imaging? The proof is in a picture: Willard S. Boyle, George E. Smith, and Michael F. Tompsett

Soon after arriving at Bell Labs in 1969, Michael F. Tompsett invented imaging charge-coupled devices (CCDs) and in the next six years grew the technology from eight-bit chips to 512 ´ 512 element arrays that matched television-screen resolution in the 1970s. That marked the start of the solid-state imaging revolution that would make today’s digital photography possible. His achievements earned him a long list of prizes including the 2010 National Medal for Technology and Innovation.

Yet when the Nobel Prize committee decided to award the 2009 Physics prize for world-changing developments in imaging and communications, Bell Labs physicists, Willard S. Boyle and George E. Smith each received a one-quarter share “for the invention of an imaging semiconductor
circuit—the CCD sensor.” Engineer Charles Kao took the other half of the prize that year for dramatic improvements in “the transmission of light in fibers for optical communication.”

While no one can know the 2009 Nobel Committee’s deliberations, the history of the development of the CCD perhaps shows why Tompsett wasn’t part of that year’s physics prize.

In the late 1960s, Bell Labs was working on several new photonic and electronic technologies that spread company resources thin. They had spent years developing a silicon-diode array vidicon tube to be the camera for the Picturephone video telephone that parent-company AT&T had introduced in 1970. For data storage, the labs were developing magnetic bubble memory, a nonvolatile thin-film technology considered very promising. Their big electronics program was metal-oxide semiconductor (MOS) circuits.

Hard-driving Bell Labs Vice President Jack Morton wanted to increase efforts on bubble memory. As Smith recounted years later in his Nobel Prize presentation, “Morton demanded that [Boyle’s semiconductor] division come up with a semiconductor device to compete with [magnetic] bubbles” for the memory application. Otherwise, he would transfer resources from semiconductors to bubbles. As a department manager reporting to Boyle, Smith was summoned to the boss’ office to figure out how.

“In a discussion lasting not much more than an hour, the basic structure of the CCD was sketched out on the blackboard, the principles of operation defined, and some preliminary ideas concerning applications were developed,” Smith said at the 2009 Nobel ceremony.

The device Boyle and Smith invented was a MOS version of a capacitor, which would confine an electric charge in a nonconductive oxide layer between two parallel metal layers. It was an electronic circuit designed as the structural analog of a magnetic bubble memory, which moved, stored, and processed electrical charges in a semiconductor like a bubble memory did magnetic fields in a magnetic material.

“This completed the basic invention,” Smith wrote. If that was the whole story, Boyle and Smith would have made their little meeting amazingly productive by any corporate standard: two Nobel prizes for a little more than an hour of brainstorming.

It was, of course, not that simple. They needed to show that it worked, and in short order Tompsett and Gil F. Amelio, who both worked for Smith, demonstrated that CCDs could quickly transfer charge with minimal loss. The results appeared in the same April 1970 issue of the Bell System Technical Journal as Boyle and Smith’s theoretical paper on CCDs.

Tompsett had come to Bell Labs in 1969 after inventing a pyroelectric camera tube for English Electric Valve. He was an imaging guy, and says he joined Bell “to get away from the tube business and into silicon.” When he first heard about the CCD, he says, “my immediate reaction was to make it into an imaging device.”

Michael F. Tompsett receives the 2010 US National Medal of Science and Technology from then-President Barack Obama. Photo credit: NMST Foundation

From his experience with camera tubes, Tompsett saw that the much smaller imaging area of the semiconductor device gave it an inherent advantage for cameras. That difference, he says, meant “the signal-to-noise ratio of CCDs was going to be much better than in existing camera tubes.” He successfully demonstrated fast and efficient charge transfer between active frames of the device, which was considered a proof of principle for CCD imaging. To convince Smith, Tompsett then had to design and build an eight-pixel CCD imager with diodes at both ends, which demonstrated both light sensing and the transfer of charge needed for imaging.

When it came time to file patents, Bell lawyers put only Boyle’s and Smith’s names on the basic patent for the CCD. They put only Tompsett’s name on the patent for CCD imaging.

Tompsett went on to manage Bell Lab’s CCD imaging group, focusing on replacing the massive tube television cameras of the time with solid-state devices. He and his colleague Ed Zimany demonstrated the first solid-state color camera by using a large prism to separate colors that were directed to three 106 ´ 128 pixel CCD arrays mounted on x-y stages. With it, Tompsett took color photos of his wife, which were also published on the cover of the January 1973 issue of Electronics magazine. Despite their low resolution, the images were a milestone.

In 1976, Tompsett demonstrated a 525-line black-and-white CCD camera, but by then AT&T had largely abandoned Picturephone, which used monochrome cameras. With other companies beginning to make CCD cameras and federal regulations then blocking AT&T from selling products it made outside the Bell System, Tompsett didn’t see any reason to continue developing cameras at Bell. Finding Bell Labs “a wonderful place for invention, with top-notch facilities and an informal, creative workplace,” Tompsett turned to other projects and stayed until 1989. He later managed a US Army laboratory and became an entrepreneur.

CCD cameras based on MOS technology opened the door for solid-state imaging systems coming into wide use in the 1980s. However, by the time of the 2009 Nobel Prize in Physics, complementary metal-oxide semiconductor (CMOS) sensors were replacing CCDs for many applications because of their cost and performance. Today the most important uses of CCD cameras are in astronomy because of their lower noise levels.

Bell Labs veterans were surprised that Tompsett was not on the Nobel list. He was the sole inventor on the patent for CCD imaging and had spent four years leading CCD development at Bell. “When you look at the Nobel Prize, it was really prompted by having these wonderful CCD imagers,” says Carlo H. Séquin, a retired University of California, Berkeley professor who worked with Tompsett at Bell. Boyle and Smith had devised the CCD circuit, but not imaging CCDs. Soon after the Nobels, the late Eugene Gordon, who worked for Boyle and was Smith’s boss, told one trade publication, “The closest Smith and Boyle got to a CCD imaging device was on a staged photo in front of one of Tompsett’s cameras.”

Boyle died in 2011 and Smith did not respond to a request for an interview. Tompsett himself had never expected anyone to get a Nobel Prize for inventing CCDs. He told one awards committee, “I pooh-poohed the idea and said it’s ridiculous. The Nobel always goes for theoretical physics, and this is engineering, so no way would they look favorably on something like that.” He’s proud of his engineering awards, and Séquin says they gave him “the credit that he so fully deserves.”

Jeff Hecht is an SPIE Member and freelancer who writes about science
and technology.