Optical fiber

Based on Wikipedia: Optical fiber

The Invisible Highway

Right now, as you read these words, pulses of light are racing beneath your feet. They're traveling through hair-thin strands of glass, carrying everything from this very text to video calls, financial transactions, and the collective chatter of billions of people. These strands—optical fibers—have quietly become the nervous system of modern civilization.

What makes this remarkable is just how simple the core idea is. Take a piece of glass, make it very pure, and shape it so that light bounces around inside rather than escaping. That's it. That's the principle that now carries over ninety percent of all intercontinental data traffic.

But simple doesn't mean obvious. It took over a century to get from "interesting physics demonstration" to "backbone of the internet." And the story involves bent glass rods in Vienna, a Nobel Prize, and a chemical engineer who figured out how to make the whole thing affordable.

How Light Gets Trapped

To understand optical fiber, you need to understand a phenomenon called total internal reflection. It's the same thing that makes a swimming pool's surface look like a mirror when you're underwater looking up at a steep angle.

Here's what happens. When light moves from one material to another—say, from water to air—it bends. This bending is called refraction, and it's why a straw looks kinked when you put it in a glass of water. The amount of bending depends on the angle at which the light hits the boundary between the two materials.

But there's a critical angle. If light hits the boundary at a steep enough angle, it doesn't pass through at all. Instead, it bounces back entirely. One hundred percent reflection, no losses.

For water, this critical angle is about forty-eight degrees from perpendicular. For flint glass, it's around thirty-nine degrees. For diamond, it's a mere twenty-four degrees—which is part of why diamonds sparkle so dramatically. The light gets trapped inside, bouncing around, only escaping through carefully cut facets.

An optical fiber exploits this same phenomenon. The fiber consists of two parts: a core of very pure glass surrounded by a cladding of slightly different glass. The cladding has a lower refractive index than the core, which is a technical way of saying light travels slightly faster through it. This difference creates the conditions for total internal reflection.

When light enters the fiber at the right angle, it hits the boundary between core and cladding and bounces back. Then it hits the opposite side and bounces again. And again. And again. The light zigzags down the length of the fiber, potentially for many kilometers, without escaping.

From Parlor Trick to Medical Tool

The physics of guiding light through transparent materials was first demonstrated in Paris in the early 1840s by Daniel Colladon and Jacques Babinet. They showed that light could follow the curve of a stream of water, bending with the liquid rather than traveling in a straight line. It was considered a delightful curiosity, the kind of thing you might demonstrate at a scientific lecture to impress an audience.

John Tyndall, the Irish physicist, did exactly that in London about twelve years later. He included light-guiding demonstrations in his public lectures and wrote about total internal reflection in an 1870 book about the nature of light. But for decades, it remained exactly that—a demonstration, a teaching tool, not a technology.

The first practical application emerged from an unlikely place: dentistry.

In the late 1800s, a team of Viennese doctors realized that bent glass rods could guide light into the body's dark cavities. Instead of trying to shine a lamp at a patient's teeth and hoping enough light bounced in, they could pipe the illumination exactly where it was needed. Dentists adopted the technique early in the twentieth century.

The next leap came from wanting to transmit not just light, but images. In the 1920s, both Clarence Hansell, a radio experimenter, and John Logie Baird, the television pioneer, independently demonstrated that images could travel through bundles of transparent tubes. The idea was sound, but the technology wasn't ready.

Heinrich Lamm, working in the 1930s, showed that bundled optical fibers could transmit images for internal medical examinations. But his work was largely forgotten, a technological dead end that would have to be rediscovered.

The Cladding Breakthrough

Early optical fibers had a fundamental problem. The core needed to be surrounded by something with a lower refractive index to create total internal reflection. The obvious choice was air—it has a much lower refractive index than glass. But air-clad fibers were fragile and impractical. Some experimenters tried oils and waxes, which worked better but were still clumsy.

In 1953, Dutch scientist Bram van Heel demonstrated something new: fibers with a transparent glass cladding. By surrounding the core glass with a different type of glass—one with a lower refractive index—he created fibers that were robust enough for practical use.

The same year, Harold Hopkins and Narinder Singh Kapany at Imperial College London achieved image transmission through bundles containing over ten thousand individual fibers. They eventually transmitted images through a bundle seventy-five centimeters long—not impressive by modern standards, but a proof of concept that changed everything.

Kapany would go on to coin the term "fiber optics" in a 1960 Scientific American article that introduced the field to a broad audience. He then wrote the first book on the subject. The name stuck.

The first major commercial application arrived in 1956: a flexible gastroscope. Developed by Basil Hirschowitz, C. Wilbur Peters, and Lawrence E. Curtiss at the University of Michigan, this device allowed doctors to look inside a patient's stomach without surgery. Curtiss solved the cladding problem by producing the first glass-clad fibers, making the device practical.

For the next decade, fiber optics meant medical instruments and specialized illumination. The idea of using fibers to transmit communication signals existed in principle, but the fibers of the era were far too lossy. Light dimmed too quickly as it traveled, making long-distance transmission impossible.

The Twenty Decibel Barrier

To understand the telecommunications challenge, you need to understand attenuation—the gradual loss of signal strength over distance. All transmission media suffer from attenuation. Radio signals fade, electrical currents dissipate as heat, and light dims.

Scientists measure this loss in decibels per kilometer. A decibel is a logarithmic unit, meaning that 20 decibels of loss represents a hundred-fold reduction in signal strength, while 40 decibels represents a ten-thousand-fold reduction. The fibers available in the mid-1960s had attenuation rates of around 1,000 decibels per kilometer—essentially opaque over any useful distance.

Charles K. Kao and George A. Hockham, working at the British company Standard Telephones and Cables, made a crucial observation in 1965. They argued that this terrible attenuation wasn't inherent to glass itself. It was caused by impurities. Remove the impurities, and fibers could theoretically achieve attenuation below 20 decibels per kilometer—a level that would make fiber optic communication practical.

Twenty decibels per kilometer became the magic number, the barrier that needed to be broken.

Kao and Hockham identified the right material: silica glass of extremely high purity. They systematically analyzed the light-loss properties and pointed the way forward. This work eventually earned Kao the Nobel Prize in Physics in 2009—recognition that came more than four decades after his original insight.

Breaking the barrier required extraordinary manufacturing precision. In 1970, researchers at Corning Glass Works—Robert D. Maurer, Donald Keck, Peter C. Schultz, and Frank Zimar—demonstrated a fiber with attenuation of just 17 decibels per kilometer. They achieved this by doping silica glass with titanium, essentially adding controlled amounts of titanium atoms to modify the glass's properties.

A few years later, they produced even better fibers using germanium dioxide as the dopant, achieving only 4 decibels per kilometer. The barrier hadn't just been broken; it had been shattered.

The Manufacturing Problem

Physics was only half the challenge. Even with fibers that could carry light efficiently, manufacturing them at scale remained prohibitively expensive and slow.

Optical fibers are made through a process called drawing. You start with a large glass cylinder called a preform—essentially a scaled-up version of the final fiber, with the core and cladding already in place. You heat one end of the preform until it softens, then pull a thin strand from the molten tip. As you pull, the strand cools and solidifies into fiber.

The precision required is extraordinary. The core of a single-mode fiber—the type used for long-distance telecommunications—is only about nine micrometers in diameter. That's roughly one-eighth the width of a human hair. The cladding brings the total diameter to about 125 micrometers, still thinner than a typical human hair.

Initially, high-quality fiber could only be drawn at about two meters per second. At that rate, manufacturing enough fiber for large-scale deployment was impossibly expensive. Copper cable, despite its limitations, remained cheaper.

The breakthrough came in 1983, when chemical engineer Thomas Mensah joined Corning. He developed techniques that increased the drawing speed to over fifty meters per second—a twenty-five-fold improvement. This dramatic acceleration made optical fiber cheaper than copper for the first time.

In 1981, General Electric demonstrated another milestone: fused quartz ingots that could be drawn into continuous strands forty kilometers long. No splices, no joints, just a single unbroken fiber stretching across twenty-five miles.

Single-Mode Versus Multi-Mode

Not all optical fibers are created equal. They fall into two main categories, and understanding the difference matters for knowing how modern networks function.

Multi-mode fibers have relatively large cores, typically around fifty micrometers in diameter. Light can take many different paths through these fibers, bouncing at various angles. Think of a hallway wide enough for people to walk side by side, taking slightly different routes.

Single-mode fibers have much smaller cores, around nine micrometers. Light can only take one path—hence the name. It's like a hallway so narrow that everyone must walk single file.

Why does this matter? Because when light takes multiple paths through a fiber, different paths have different lengths. Some rays bounce more frequently, traveling a longer total distance. This means that a short pulse of light entering the fiber gets smeared out over time as different rays arrive at slightly different moments.

This smearing, called modal dispersion, limits how fast you can send data. If pulses spread out enough to overlap, you can't tell one pulse from the next. Multi-mode fibers work fine for short distances—inside a building, for example—but become problematic over longer spans.

Single-mode fibers avoid this problem entirely. With only one path possible, there's no dispersion from multiple modes. You can send data faster and farther. Almost all long-distance fiber links use single-mode fiber, including anything spanning more than about a kilometer.

The tradeoff is that single-mode fiber is harder to work with. The tiny core means precise alignment is critical when connecting fibers. But for the backbone of the internet, where distances are measured in thousands of kilometers, single-mode is the only practical choice.

Connecting Fibers Together

Joining optical fibers is far more complicated than splicing electrical wires. With copper wire, you can strip the insulation, twist the conductors together, and you're done. With fiber, you need to align two glass cores, each thinner than a human hair, with sub-micrometer precision.

For permanent connections, the standard technique is fusion splicing. The process involves carefully cleaving both fiber ends to create perfectly flat, perpendicular faces. The ends are then positioned precisely using specialized equipment that can measure alignment to fractions of a micrometer. Finally, an electric arc melts the glass, fusing the two fibers into a continuous piece.

A good fusion splice has very low loss—typically less than 0.1 decibels. That's a loss of only about two percent, barely detectable.

Mechanical splices offer another option. Instead of melting the fibers together, these devices hold the cleaved ends in precise alignment using mechanical force and index-matching gel to fill any tiny gap. Mechanical splices are faster to make and don't require expensive fusion equipment, but they typically have higher loss and are less durable.

For connections that need to be made and broken repeatedly, specialized optical fiber connectors are used. You've probably seen these if you've connected audio equipment—the red-lit ports on home theater receivers use an optical connection standard called TOSLINK. These connectors include ferrules that hold the fiber in precise position, allowing reasonably low-loss connections without permanent splicing.

The Amplifier Revolution

Even the purest fiber attenuates light over distance. For very long links, the signal eventually becomes too weak to detect. The original solution was repeaters: devices that detect the optical signal, convert it to electricity, amplify the electrical signal, and convert it back to light.

These optical-electrical-optical repeaters worked, but they were expensive and complicated. Each repeater had to handle the specific data format being transmitted, making upgrades difficult. And placing repeaters every few tens of kilometers across an ocean was an engineering nightmare.

The game-changer was the erbium-doped fiber amplifier, developed in 1986 and 1987 by two teams working independently—one led by David N. Payne at the University of Southampton, the other by Emmanuel Desurvire at Bell Labs.

Erbium is a rare earth element. When added to glass fiber in small quantities, it has a remarkable property: if you pump energy into the erbium atoms using light at one wavelength, they can amplify light passing through at a different wavelength—specifically, the wavelengths used for telecommunications.

This means you can amplify the optical signal directly, without converting it to electricity. The amplifier doesn't care what data format is being transmitted; it simply makes the light stronger. This dramatically reduced the cost of long-distance fiber systems and enabled technologies like wavelength-division multiplexing.

Many Colors, Many Channels

Wavelength-division multiplexing, usually abbreviated W-D-M, is one of those technologies that sounds complicated but is conceptually simple. It's the optical equivalent of radio stations using different frequencies.

Light, like radio, exists across a spectrum of wavelengths. What we perceive as red light has a longer wavelength than blue light. Fiber optic systems typically use infrared light—wavelengths longer than visible red—but the principle is the same.

With wavelength-division multiplexing, you send multiple signals through the same fiber simultaneously, each using a different wavelength of light. At the receiving end, you separate the wavelengths and recover each signal individually. It's like having multiple radio stations broadcasting at once, but through a single cable.

Commercial systems commonly use eighty or more separate wavelengths in a single fiber. If each wavelength carries ten gigabits per second of data—a typical rate—that's eight hundred gigabits per second through a single strand of glass thinner than a human hair.

The erbium-doped fiber amplifier made this practical because it amplifies all the wavelengths simultaneously. Without it, you'd need a separate amplifier for each channel, making the whole scheme economically unviable.

Beyond Total Internal Reflection

In 1991, a new type of fiber emerged from the field of photonic crystals: photonic-crystal fiber. These fibers guide light not through total internal reflection but through diffraction from a periodic structure—essentially a pattern of microscopic holes running along the fiber's length.

The first photonic crystal fibers became commercially available in 2000. They offer capabilities that conventional fibers can't match. They can carry higher power levels without damage. Their properties can be engineered by adjusting the pattern of holes. And remarkably, they can have hollow cores—air rather than glass in the center.

A hollow-core fiber sounds impossible. How can you guide light through empty space? The photonic crystal structure creates a kind of optical cage that keeps the light confined even without a solid material to reflect it. This opens applications in sensing, high-power delivery, and ultrafast pulse transmission.

Sensing Instead of Signaling

Optical fibers don't just carry communications. They also make excellent sensors.

The basic idea is straightforward: if something changes the fiber's properties, it changes the light passing through. Measure the light, and you've measured whatever changed the fiber. Temperature, pressure, strain, vibration—all of these can be detected.

The simplest fiber sensors measure intensity. If bending the fiber reduces how much light gets through, you can detect bending by measuring light output. More sophisticated sensors measure changes in the light's phase, polarization, or wavelength, providing greater sensitivity and specificity.

One particularly powerful technique is distributed acoustic sensing. A single optical fiber, stretched over many kilometers, can detect vibrations along its entire length with remarkable precision. This makes it possible to monitor pipelines for leaks, fences for intrusions, or railway tracks for approaching trains—all from a single instrument connected to one end of the fiber.

The opposite extreme is also possible: sensors so small they fit on the fiber's tip and can be inserted into blood vessels through a needle. These microscopic sensors enable measurements in places no other technology can reach.

Fiber optic gyroscopes represent another sensing application. They detect rotation using a phenomenon called the Sagnac effect: light traveling around a loop in the direction of rotation takes slightly longer than light traveling against the rotation. By measuring this tiny time difference with extreme precision, the gyroscope detects rotational motion with no moving parts.

Powering Devices With Light

Here's something counterintuitive: you can transmit power through optical fiber. It's not efficient compared to copper wire, but efficiency isn't always the point.

The technique uses a photovoltaic cell at the receiving end—similar to a solar cell—to convert light back into electricity. Why would anyone do this instead of just running a wire?

Consider a magnetic resonance imaging machine, the kind used in medical diagnostics. These devices produce extremely strong magnetic fields. Metal conductors near the machine can distort the field, ruining the image. Or worse, they can heat up from induced currents, potentially causing burns. Fiber optic power delivery avoids both problems because glass doesn't interact with magnetic fields.

Similar logic applies to high-voltage equipment, where fiber can safely bring power to monitoring devices that would be damaged or disrupted by direct electrical connection to the high-voltage system.

Light Pipes and Endoscopes

Not every fiber optic application involves telecommunications or sensing. Some fibers simply carry light where it's needed.

Medical endoscopes bundle thousands of fibers together in a coherent arrangement—meaning the fibers maintain their relative positions from one end to the other. Put a lens on each end, and you have a long, flexible imaging device. Light goes in through some fibers, illuminating whatever is at the far end. The image comes back through other fibers to be viewed by the physician.

This enables minimally invasive surgery: procedures performed through tiny incisions rather than large openings. The surgeon watches a screen showing the endoscope's view while manipulating instruments through the same small incision.

Industrial versions called fiberscopes or borescopes inspect machinery interiors, jet engine turbines, and anything else that can't be examined directly. You thread the flexible tip into the confined space, and the image travels back through the fiber bundle.

Even more creative applications exist. Some buildings use fiber optic systems to pipe sunlight from the roof to interior spaces, providing natural illumination without windows. Decorative applications range from fiber optic lamps to artificial Christmas trees with glowing tips.

There's even a building material called LiTraCon—light-transmitting concrete—that incorporates optical fibers. The concrete functions structurally like normal concrete, but light passes through via the embedded fibers. The effect is walls that appear solid but allow ghostly images of whatever is on the other side.

The Battlefield Goes Optical

In March 2024, a new application emerged from the ongoing conflict in Ukraine: fiber optic drones.

Modern warfare makes extensive use of electronic warfare—systems designed to jam or interfere with enemy communications and control systems. Radio-controlled drones are vulnerable to this jamming. Disrupt the radio link, and the drone loses contact with its operator.

Fiber optic drones solve this problem by trailing a thin optical fiber as they fly. Commands travel down the fiber to the drone, and video feeds come back up. Because the signals are light rather than radio waves, electromagnetic jamming has no effect. The drone remains under positive control regardless of what electronic warfare systems the enemy deploys.

It's a reminder that optical fiber, for all its association with civilian telecommunications, has capabilities that extend far beyond carrying internet traffic.

The Modern Network

Today's long-haul fiber connections span continents and cross oceans with remarkably few interruptions. Repeater spacing—the distance between amplifiers—typically ranges from seventy to one hundred fifty kilometers. Undersea cables stretch thousands of kilometers with amplifiers spaced along their length.

Attenuation in modern fiber is far below what Kao and Hockham dreamed of achieving. Current fibers achieve losses around 0.2 decibels per kilometer, meaning a signal travels five kilometers before losing half its strength. Compare this to the thousand decibels per kilometer of 1960s fibers, and the progress becomes stark.

Speed continues to increase. Systems transmitting ten or forty gigabits per second per wavelength are common in deployed infrastructure. Laboratory demonstrations have pushed into the petabit range—millions of gigabits per second through a single fiber. The capacity of optical fiber, it seems, still has room to grow.

Optical fiber has become so central to modern communications that it's easy to forget how recently it was merely theoretical. The first commercial fiber optic link was installed in Turin, Italy, in 1977—less than fifty years ago. The person who made it economical, Thomas Mensah, was still at Corning in 1983.

The technology that carries nearly all of the world's data traffic is younger than most of the people reading about it. What began as a physics demonstration in Paris, showing that light could follow a curved path through water, now forms the invisible infrastructure on which civilization increasingly depends.