How Charles Kao Beat Bell Labs to the Fiber-Optic Revolution

Fifty years ago this month, a 32-year-old Chinese-born research engineer named Charles Kao published a milestone paper that set off the entire field of fiber-optic communications and eventually earned him a share of the 2009 Nobel Prize in Physics. The story of how that 1966 paper came to be is a wonderful example of how the key to a big technological breakthrough can come down to asking the right question.

It began in the early 1960’s. The invention of the laser had generated a wave of interest in using light to transmit information at far greater rates than is possible with the lower frequencies of radio. But experiments sending laser beams through the air quickly showed that clouds and precipitation got in the way.

Bell Labs decided laser signals should go through hollow metal pipes buried in the ground. This was a logical step; the lab was already developing hollow waveguides to transmit high-frequency microwaves. And it seemed obvious that clear air should be far more transparent than any solid could be.

But Antoni Karbowiak, Kao's then-boss at Standard Telecommunications Laboratories—an IT&T subsidiary in Harlow, England—took a different approach. Karbowiak was trying to make an optical analog of solid plastic waveguides, which could be bent to make connections between pieces of microwave equipment. He and his colleagues had heard that bundles of glass fibers were being developed for medical imaging inside the body. They had also heard that Eli Snitzer at American Optical, in Massachusetts, had shown that a very thin optical fiber could be used to confine light to a single path, avoiding the degradation that happens when a signal is divided among multiple paths with different distances and arrival times. Then Karbowiak left for Australia to accept a professorship, leaving Kao in charge of the optical communications project. 

Kao went to Harold Rawson, a specialist in glass technology at the University of Sheffield in Britain, and asked a critical question: how transparent can glass possibly be made? In particular, could the attenuation of light be reduced to 20 decibels per kilometer, low enough for signals to travel several kilometers before needing to be amplified? Rawson thought so, because most absorption in glass is due to impurities. No one had put much effort into driving down the concentration of such impurities because there was no reason to try sending light through glass further than the few meters needed for short-range medical purposes. That was one of several key points that Kao and his assistant George Hockham made in their 1966 paper outlining how a fiber-optic communication system could work.

Kao and Hockham weren’t the only ones studying the prospects of fiber around that time. Rudolf Kompfner, a top manager famed for stimulating a creative generation at Bell Labs, also considered the feasibility of sending laser beams through glass fibers. But when he did, he asked a different question: how transparent were the clearest glass fibers available today? That answer was on the order of a thousand decibels per kilometer, which means that a kilometer of glass would attenuate light by a factor of 10¹⁰⁰, effectively consuming every last photon. It was no wonder Bell continued with its hollow optical waveguides despite the fact that they required an elaborate system of lenses to prevent too much loss from light bouncing off their metal walls.

The British Defense Ministry and the British Post Office's telecommunications branch were interested in Kao's proposal. But Kao also wanted to persuade others that fiber-optic communications was feasible. To bolster his case, he looked past the glass fibers then in existence for the purest glass sample he could find: fused silica made by a vapor deposition process. Corning Glass Works, based in Corning, N.Y., had developed the material to withstand high temperatures, but had not studied its optical properties in detail. Kao's exacting measurements showed fused silica was even more transparent than Rawson had predicted.

The next challenge was drawing fused silica into light-guiding fibers that retained that vital clarity. In 1970, a group at Corning consisting of Robert Maurer, Donald Keck, Peter Schultz, and Frank Zimar, made light-guiding fibers from fused silica with attenuation of only 17 dB/km, below Kao’s original target. Bell Labs and others were stunned. In 1972, Corning made a stronger fiber with a loss of only 4 dB/km, and by 1979, glass fiber loss had dropped as low as 0.2 dB/km. Fiber-optic communications became a dramatic success and the backbone of the global telecommunications system.

Bell Labs got on the bandwagon in the early 1970’s. Their plan had been to introduce buried microwave waveguides, which had then been in development for some 20 years, and to eventually replace those with hollow optical waveguides. But fiber offered such high capacity and was so easy to install that it blew away microwave pipes before any were put into commercial service. If those bulky, hollow tubes could have been made to work, they would have been hugely expensive to install and could never have matched the capacity of today's fiber systems.

What made fiber’s success possible was Kao's question. He asked what was possible to do, not what had been done. That was a crucial difference, because nobody had tried to make glass so transparent before. Nobody had needed to send light so far through glass. By good fortune Corning had developed fused silica for its strength and thermal properties, and as a result, the company had a head start on making ultra-transparent fiber. So think carefully before you ask what's been done or what's possible.

For more, see the video below of Kao, a Life Fellow of IEEE, working in the lab and this oral history conducted the IEEE History Center.

Jeff Hecht is the author of City of Light: The Story of Fiber Optics (Oxford University Press).