Published by: Research & Development Department, Technologie Optic Inc., September 2025

Introduction

For centuries, humans have sought ways to use light for communication – from beacon fires and semaphore telegraphs to modern laser links. In free space, however, light energy spreads out and diminishes rapidly with distance. This follows the inverse-square law, which states that the intensity of light drops proportional to the square of the distance from the source. In practical terms, a receiver far from a light source collects only a tiny fraction of the emitted light, leading to weak signals.

Figure 1 illustrates how light intensity decreases as distance increases. To overcome this geometric dilution of light, engineers employ waveguides – structures that confine and direct light. An optical fiber is essentially a transparent waveguide that channels light in a single direction, much like a pipe carries water without spilling. By guiding the light, optical fibers prevent the dispersion of energy into space, allowing efficient transmission of information over long distances without the severe loss that a free-spreading beam would incur. This principle enabled the development of today’s high-capacity fiber-optic communication networks that span the globe.

Silica Optical Fibers

Harnessing light for long-distance communication became feasible with the invention of the silica optical fiber in the late 20th century. While rudimentary light-guiding concepts date back to John Tyndall’s 1870s water-fountain demonstration of total internal reflection, modern fiber optics truly began in the 1960s. In 1966, Dr. Charles K. Kao and George Hockham published a landmark paper proposing that purified silica glass fibers could be used to carry signals over long distances [1]. At the time, glass fibers had extremely high loss (over 1,000 dB/km), meaning light could only travel a few meters before fading out. Kao calculated that if fiber attenuation could be reduced below about 20 dB/km, optical communication over kilometers would be possible. This bold vision earned Kao the nickname “Father of Fiber Optics” (and a share of the 2009 Nobel Prize in Physics). It spurred a worldwide race to create ultra-pure glass fibers.

The breakthrough came in 1970 when researchers at Corning Glass Works (Robert Maurer, Donald Keck, Peter Schultz, and Frank Zimar, see Figure 2) achieved the first low-loss optical fiber. Their silica fiber had an attenuation of about 17 dB/km at red light wavelengths– a staggering improvement, though still higher than Kao’s target. Progress was swift: by 1972 Corning had fibers with only ~4 dB/km loss, and by 1979 losses dropped to ~0.2 dB/km in the optimal infrared wavelengths.

For context, 0.2 dB/km means over 50% of light remains after 1 km of fiber, enabling fiber links of tens of kilometers with periodic amplification. This achievement “lit up” the field of fiber-optic communications and laid the foundation for the global internet. Today, silica fiber forms the backbone of telecommunications, reliably carrying data, phone, and video signals at near light-speed across oceans and continents.

Fundamentals of Silica Optical Fibers

The magic of optical fibers lies in a simple but powerful optical principle: total internal reflection (TIR). As illustrated in Figure 3, a fiber consists of a transparent silica core surrounded by a lower-index cladding. The cross-sectional view highlights this layered structure, while the side view shows how light rays are repeatedly reflected within the core, allowing signals to propagate over long distances with minimal loss. When light traveling in the core reaches the core–cladding boundary at a shallow angle, it is reflected back into the core rather than refracted out. If the incident angle is below a certain critical angle, all the light is reflected internally – this is total internal reflection. In essence, the fiber’s cladding acts as a mirror that keeps the light trapped within the core. Once light is injected into the core at an appropriate angle (within the fiber’s acceptance cone defined by its numerical aperture), it zig-zags down the fiber, reflecting off the internal walls and maintaining its path down the length of the fiber. By this mechanism, fibers can guide light around bends and over great distances with minimal loss. The glass itself is ultra-transparent pure silica, so little light is absorbed as it travels (modern fibers absorb only a few percent of light per kilometer). A protective buffer coating (usually acrylic polymer) surrounds the cladding to provide mechanical strength and keep moisture or impurities out.

Single-Mode and Multi-Mode Fibers

Not all optical fibers are the same – they come in different types optimized for various applications. The two primary categories are single-mode fiber (SMF) and multimode fiber (MMF). The distinction lies in the core diameter and how light travels in the fiber:

Single-Mode Fiber: As its name indicates, a single-mode fiber contains a very narrow core, typically around 8–10 µm in diameter (see Figure 3). This small core only allows one path (mode) for the light to propagate. Essentially, the light travels in a single ray-like mode straight down the fiber. By eliminating multiple paths, SMF avoids the problem of different light rays arriving at different times. This gives highest bandwidth and longest reach – ideal for long-haul communication links, high-speed data networks, and undersea cables. Typical single-mode fibers are designed for wavelengths around 1310 nm and 1550 nm, where silica loss is lowest. Indeed, most telecom systems use single-mode fiber with infrared lasers at 1550 nm (and 1310 nm for some shorter links). The narrow core of SMF requires precise laser sources and alignment, making the transmitters slightly more expensive, but it repays with virtually unlimited distance and data rate (tens of terabits per second over hundreds of kilometers with amplification).

Multimode Fiber: A multimode fiber has a larger core—typically 50 or 62.5 µm—which supports multiple light paths (modes), as depicted in Figure 3. LEDs or lower-cost VCSEL lasers (typically at 850 nm wavelength) can easily excite multimode fibers. MMF is used mostly for short distance links such as within buildings or data centers (up to a few hundred meters). Because multiple modes travel at slightly different speeds, a short light pulse tends to spread out (modal dispersion) over distance, limiting the data rate and distance compared to single-mode. Multimode fiber is usually optimized for wavelengths of 850 nm and 1300 nm (the so-called first and second telecom windows for older fiber). It’s perfect for economical connections in LANs, factory networks, or audio/visual equipment over tens to hundreds of meters. The trade-off is higher attenuation (~3 dB/km at 850 nm, higher than SMF at 1550 nm) and significantly lower bandwidth-distance product, but the equipment (transceivers, connectors) can be cheaper and more tolerant of slight misalignment due to the wider core. In summary, single-mode = long distance, high bandwidth; multimode = short distance, lower cost.

Transmission Windows in Silica Fiber

Silica optical fibers have certain preferred wavelength windows where they exhibit minimal signal loss and distortion. This is because the attenuation in glass isn’t uniform across all colors of light – it depends on material absorption and scattering characteristics. Telecom fibers operate in the near-infrared range, roughly 850 nm to 1600 nm, where silica is most transparent. There are several standard bands (or “windows”) in this range:

• First window (850 nm band): Historically used with early multimode fibers and LED sources. Attenuation here is higher (~2–3 dB/km), and scattering is more significant at shorter wavelengths. Today, 850 nm remains important for short multimode links (e.g., in data centers), often using VCSEL lasers, but it’s not used for long-haul links due to loss. (This corresponds to the upper visible/near-IR region and is sometimes just called the “850 nm band”.)
• O-band (1260–1360 nm): The Original band used in optical communications. Early single-mode fiber systems in the 1980s used ~1310 nm because fiber dispersion is very low there (pulses don’t spread out much) and the material could be made with reasonably low loss. The O-band was attractive since fibers in the 1970s had their minimum loss near 1300 nm. Attenuation in the O-band is typically around 0.4–0.5 dB/km in modern fibers – good, but not the absolute best.
• E-band (1360–1460 nm): The Extended band. This region was initially avoided because early fibers had a high loss spike around 1383 nm due to absorption by traces of OH⁻ (water) in the glass. Improved manufacturing (dehydration techniques) later reduced this “water peak,” allowing the E-band to be used in some applications. However, many legacy fibers still have high loss in this range, so E-band is less commonly utilized in practice.
• S-band (1460–1530 nm): The Short wavelength band just below C-band. Fiber attenuation here is slightly higher than the absolute minimum, but still quite low, and S-band can be used for additional capacity in some systems. Passive optical networks (like fiber-to-the-home systems) often use around 1490 nm (in S-band) for downstream signals. There is growing interest in using S-band alongside C-band to expand fiber bandwidth (with new amplifier technologies).
• C-band (1530–1565 nm): The Conventional band, and the most widely used window for long-distance communication. This is where silica fiber’s attenuation hits its all-time low – around 0.2 dB/km, thanks to minimal Rayleigh scattering and very low intrinsic absorption. The C-band is the workhorse of optical telecommunications; it’s used in undersea cables and backbone networks. Erbium-doped fiber amplifiers (EDFAs), a critical fiber-optic technology, are also optimized for the C-band, enabling easy amplification of signals in this range. (1550 nm, the center of C-band, is often considered the “third window” of fiber optics and the optimal wavelength for loss.)
• L-band (1565–1625 nm): The Long wavelength band. Silica’s loss remains low in L-band – only slightly higher than C-band. Many modern systems have extended into L-band to roughly double the available spectrum, especially for high-capacity dense WDM (wavelength-division multiplexing) systems. EDFAs can be made for L-band as well. Using C+L bands together is now common in submarine cables and long-haul networks to carry more channels of data.
• U-band (1625–1675 nm): The Ultra-long wavelength band, at the edge of silica’s low-loss range. Here fiber attenuation starts rising due to infrared absorption. U-band isn’t typically used for data channels, but it is sometimes reserved for fiber monitoring and network maintenance signals (since it’s beyond the range of normal traffic, one can send a test signal to check fiber health).

Figure 4 shows a representative attenuation curve for silica fiber versus wavelength, highlighting these telecom bands and their loss characteristics. By choosing wavelengths in these optimal windows, engineers minimize signal attenuation and dispersion, enabling data to travel tens or hundreds of kilometers between repeaters. Modern fiber systems often operate around 1310 nm (for shorter links or where low dispersion is needed) and 1550 nm (for longest links and dense wavelength multiplexing). The development of low-loss fiber “windows” is one of the reasons optical networks can carry massive amounts of information over transcontinental distances with relatively few amplifiers.

Technologie Optic.ca, a Canadian leader in fiber-optic innovation, actively develops and supports fiber systems across the full telecom spectrum—from the O-band to the U-band. Their product lines include transceivers and direct attach cables optimized for wavelengths ranging from 850 nm to 1675 nm, enabling robust performance in both short-range and long-haul applications.

Challenges and Limitations of Silica Fiber

Conventional solid-core silica fibers have enabled incredible advances in communications, but they also come with challenges and limitations:

Attenuation and Bandwidth Limits: Despite being extremely transparent, silica fiber isn’t perfectly lossless. Even at 1550 nm with ~0.2 dB/km loss, a signal will lose about half its power every 15 km. Over long distances, fibers need periodic optical amplifiers or repeaters. Moreover, fiber bandwidth is high but not infinite – effects like chromatic dispersion (wavelength-dependent speed) can blur signals over long distances, requiring dispersion compensation or sophisticated modulation formats. The defined telecom windows exist because outside those ranges, loss climbs steeply (e.g., due to ultraviolet absorption at shorter wavelengths and material vibrations absorbing light at longer wavelengths). Engineers must carefully select lasers and fiber types to optimize performance for the application, balancing attenuation and dispersion.

Physical Handling and Contamination: Silica fibers are hair-thin (125 µm outer diameter for the glass) and need careful handling. When fibers are spliced or connectors are plugged, the glass surfaces must be extremely clean. Any tiny speck of dust or dirt on a fiber connector end-face can cause significant optical loss and signal reflections. A contaminated connector not only degrades the signal, it can even damage the fiber or receiver – for instance, dirt on a fiber end can get burned onto the glass by the intense light, causing permanent attenuation. Thus, fiber-optic cables require clean, polished connections and often protective caps when disconnected. Technicians are trained to meticulously inspect and clean connectors with specialized tools, as even microscopic debris can disrupt a 40 Gb/s or 100 Gb/s data stream. Additionally, while fibers can bend around corners, bending too tightly can cause light to leak out (bending loss) or even break the fiber, so minimum bend radius guidelines must be observed (typically no sharper than a few centimeters radius for standard fiber).

The research and development team at Technologie Optic.ca is actively working on contamination challenges in silica fibers by applying cutting-edge scientific techniques. Their approach involves exploring both the structural composition and the fundamental nature of various contaminants to understand how these impurities alter the crystalline structure of the fiber. Following this analysis, the team aims to develop methods to either eliminate these contaminants or mitigate their impact on the performance of silica-based optical fibers.

Nonlinear Optical Effects: When optical power in a fiber is pushed very high or when many channels are WDM-multiplexed, the fiber’s nonlinearities come into play. In a fiber, light is tightly confined to a tiny cross-section, which means the optical intensity (power per area) can become quite large even at moderate power levels. Over tens of kilometers, these intensities induce nonlinear interactions in the glass. The most common is the Kerr effect, where the refractive index of silica increases slightly with light intensity. This can cause phenomena like self-phase modulation, cross-phase modulation (between different wavelength channels), and four-wave mixing, all of which can distort the signal or create interference between channels. Other nonlinear effects include Raman scattering and Brillouin scattering, which can transfer light to different frequencies and cause loss or noise. The Kerr nonlinear response of silica imposes a fundamental limit on how much information you can send through a single fiber – often termed the “nonlinear Shannon limit” in fiber communications. To manage these effects, systems either limit the power per channel, use advanced signal processing to compensate distortion, or even new fiber designs to mitigate nonlinearity. Nonlinearities are not a factor at low power (e.g., short fiber runs or slower links), but in long-haul high-data-rate cables carrying dozens of intense wavelength channels, they become a critical design concern.

Latency (Speed of Light in Fiber): Although we often tout communications happening at "the speed of light", in fiber that actually means the speed of light in glass, which is a bit slower than in a vacuum. Light in silica travels at about 2/3 the speed of light in air/vacuum due to the refractive index (~1.5) of glass. In numbers, light goes ~300,000 km/s in vacuum, but roughly 200,000 km/s in fiber. This means there is an inherent latency of about 5 microseconds per kilometer of fiber. Over transoceanic distances (say 10,000 km), fiber links incur on the order of 50 milliseconds of one-way delay just from propagation time. In most applications this slight delay is negligible (the advantages of fiber far outweigh it), but in certain high-frequency trading or supercomputing contexts, even these milliseconds matter. Efforts to reduce latency include finding straighter routes for fibers, and in cutting-edge cases, exploring new fiber types (as we’ll discuss next) that allow light to travel faster.

The Rise of Hollow-Core Fiber

To address the inherent limitations of conventional silica fibers, researchers have introduced hollow-core fibers (HCFs) – an advanced class of optical fibers in which light travels through an air-filled (or vacuum) core instead of solid glass. Their operation relies on precisely engineered cladding structures, such as photonic bandgap or anti-resonant designs, which confine and guide light within the hollow center. Because the optical field interacts predominantly with air rather than silica, HCFs exhibit several distinct performance advantages. As a forward-looking company, Technologie Optic.ca is actively involved in the development of next-generation fiber technologies, including hollow-core fibers. Their research and product offerings reflect a commitment to pushing the boundaries of optical transmission, from conventional silica fibers to advanced air-guided designs. For an in-depth discussion of HCF technology, please refer to our dedicated Hollow-Core Fiber paper available in the Knowledge Base section of Technologie Optic.ca website.

Technologie Optic.ca’s engineering team is actively designing advanced communication systems that operate across a wide range of wavelengths—from short-range multimode links to long-distance single-mode networks. Their work spans the full telecom spectrum, including the O-band through the U-band, ensuring compatibility with modern infrastructure and emerging technologies. For more information or to connect with our experts, please visit https://www.optic.ca/ .

Reference