Hole-way to Heaven
Hollow core fiber could lead the way to much lower cost networks
By Annie Lindstrom
from the June 4, 2001 issue of Broadband Week
Years ago, breakfast cereal shaped like a honeycomb was created. Kids loved it, but the new shape didn't cause a cereal revolution. Today, optical fiber entrepreneurs and incumbent manufacturers are giving the honeycomb a whirl in fiber design. Could it bring about revolutionary changes in the fiber optics world?
Philip Russell, chief technology officer of BlazePhotonics certainly thinks so. He's spent the past 10 years designing so-called holey fiber. And about six years ago, while on staff at England's University of Bath, he and a team of fellow researchers built a prototype that proved that holey fiber could transmit light through a hollow, air-filled core.
In lay terms it's called holey fiber because that's exactly what it looks like. To industry insiders it's called photonic crystal fiber (PCF), because it uses a two-dimensional crystal structure to keep light inside the fiber. Whatever you call it, it seems like a crazy idea. Using holes to keep light inside something simply doesn't jibe with most people's idea of what holes are supposed to do.
But if the lattice of holes is constructed just so and the holes are sized properly, they form what's called a photonic band gap (PBG). Ironically, the PBG keeps certain wavelengths confined more effectively than the total internal reflection used in today's conventional fiber.
To understand how the PBG keeps light from escaping the hollow core, take a look at the glass door of your microwave next time you use it to heat something up. The holey screen behind the glass keeps microwaves from escaping through the glass in much the same way the PBG keeps certain wavelengths from escaping from the core of PCF.
"It's like reinventing fiber completely," Russell explains.
Dear Prudence
The big question among those who are keeping a close eye on PCF is, does fiber really need to be reinvented?
"Today's glass core fibers are so fabulously good that they are going to be really hard to beat," says Doug Allan, research associate for Corning, which is studying PCF and some of its own developments in its lab. "But if we could get past the practicalities standing in our way and make a dream fiber that has an air core, it would have possibilities that could not be matched by putting light through glass."
Filling the hollow fiber core with gas, a vacuum or other materials to see the impact on light is another possibility researchers are thinking about, Allan adds.
Beaming laser light through air has many advantages over beaming it through erbium-doped silica, according to Russell. In the first place, using air as the transport medium almost entirely eliminates optical non-linearities, dispersion and Fresnel reflections. In the second place, the fiber can be made of pure silica, which is cheaper than erbium doped fiber.
If (and it's a big if) BlazePhotonics and other companies working to produce PCF can overcome the obstacles standing in the way of commercially manufacturing PCF, it's entirely possible that service providers could use it to deploy an undersea or cross-country cable without any repeaters. The cost savings potential of such a feat is enormous. But so are the hurdles.
Reality Bytes
To make the honeycomb-esque fiber, the form from which the fiber is pulled (think of a test tube of glass heated and pulled from a tower hundreds of feet high) will consist of multiple, capillary-like, hollow silica tubes stacked on top of one another. Today's fiber is pulled from a solid silica tube with an erbium-doped core.
To make hollow core fiber, the glass tube capillaries in the center of the form are removed. The glass is heated and pulled like conventional fiber, according to Russell. One issue, among many others, standing in the way of a continuous manufacturing process is finding a way to keep the glass clean during the manufacturing process. The fiber Russell is producing today is still too lossy for commercial use, he adds. BlazePhotonics was formed as a spinoff of the University of Bath's research team in order to raise the money--$9 million in its first round of funding--and spend the time needed to solve these issues, he explains.
"We almost take for granted the transparency of what you can buy off the shelf today," says Allan. "Many years of work went into learning the processes we use now."
It's all very exciting, but those who have their hopes pinned on holey fiber need to remain patient. David DiGiovanni, director of optical fiber research for Bell Labs, points out that it took 12 years for fiber manufacturers to figure out how to make low-loss silica fiber. Bell Labs is doing "fundamental research" on holey fiber, he adds.
What Is It Good For?
Besides hollow core fiber, BlazePhotonics is developing two other types of holey fiber. Endlessly Single Mode Fiber can be constructed with a large core that measures 20 microns across. It can handle many times more power than conventional single-mode fiber, according to Russell. That may come in handy as vendors develop systems that can send more and more wavelengths into a fiber. Each wavelength is associated with a certain amount of power.
The second type, Polarization Maintaining Fiber, has a unique characteristic with its shorter beat lengths, which make it a particularly useful medium for applications in which preserving the polarization state of light is important.
If PCF becomes commercially viable, it likely would be used as long-haul fiber and in highly non-linear components, says Russell. He is hesitant to be more specific, though. "I'm not free to discuss those things. We don't want to give away any secrets," Russell says.
Amplifying fibers with large effective areas could be among the first PCF products to come to market, says Pierre Sansonetti, director of Alcatel's Optical Fiber Component Research Unit. Alcatel has been working on holey fiber at the research level for about two years, he adds.
Russell guesses that PCF-based products will be used in real networks within the next two to three years as long haul fiber or as some sort of module in an all-optical switch.
BlazePhotonics is not the only company working furiously to bring hollow core fiber to the commercial market. Massachusetts Institute of Technology (MIT) spinoff Omni-Guide Communications is pursuing the same end using a one-dimensional photonic crystal, according to Uri Kolodny, the company's co-founder and director of marketing.
Instead of a honeycomb design, Omni-Guide's fiber features a hollow core with alternating layers of a highly reflective, omnidirectional, dielectric mirror material surrounding it. The product is based on the work that company chief technology officer and co-founder Yoel Fink did while earning his PhD at MIT.
Unlike BlazePhotonics, Omni-Guide has yet to produce any product. Nevertheless, the company raised $4 million in financing in August 2000 and Kolodny believes his fiber will be easier to manufacture and confine light better than 2D, honeycomb style PCF.
"We are predicting very low losses, more than an order of magnitude lower than the theoretical limit in silica fiber," Kolodny says. "We have simulated the system in a computer, but we have not manufactured the fiber yet. That's the name of the game and our current mission is to demonstrate a continuous manufacturing process."
Omni-Guide "hopes" to have drawn fiber using its structure by year's end. Although there still is a lot of risk involved in the venture and a lot of work left to be done, service providers and component manufacturers should be pretty excited about hollow core fiber, he adds.
"There are tremendous possibilities to use this fiber to simplify the design of optical networks and do things that are very difficult or expensive today," Kolodny says.
Win/Win
If BlazePhotonics and Omni-Guide both fail to find a way to produce their next-generation fibers, it won't be the first time a good fiber idea has fizzled, says DiGiovanni. In the mid-80s a group from France invented a fluoride-based fiber, which promised on paper to deliver 1/100 of the loss of silica fiber. Fifteen years and a $1 billion later no one has been able to make a fluoride fiber that spans longer than a hundred meters.
"The payoff would have been astronomical, but the hurdles were too significant," DiGiovanni says.
If PCF suffers the same fate, all will not be lost. In their PCF research, fiber makers are putting photons, which like electrons move in wave patterns, through the same drills they put electrons through to develop the transistor and computer chip.
"It's the wave property that is being manipulated by the periodic structure of the crystal, says Allan. "What people are doing is going back into all the technology and calculations associated with the semiconductor industry and trying to see if what holds true for electrons also holds true for photons. In the process, we are learning the basics of how light moves through different structures, which is kind of an end in itself."
|
