Light Emitting Diodes (LEDs) in Optical Communications

Understanding the PN junction, semiconductor doping, light emission, wavelength selection, and the properties that limit LEDs in optical-fiber communication.

The PN Junction

A PN junction is formed by bringing together P-type and N-type semiconductor material. To understand what happens at the junction, it helps to first understand how P-type and N-type material is created in a silicon crystal.

Although silicon is not the principal material used in most communication LEDs, silicon provides a useful starting point because its crystal structure and doping behavior clearly illustrate the principles of semiconductor operation.

Silicon has an atomic number of 14, meaning that a neutral silicon atom contains 14 electrons. These electrons occupy different energy levels, with four electrons in the outermost, or valence, shell. Silicon is therefore tetravalent and can form four covalent bonds with neighboring atoms.

Silicon occurs naturally primarily in the form of silica, or silicon dioxide (SiO₂). While silica is the starting material for manufacturing the glass used in optical fiber, electronics requires elemental silicon. Oxygen is removed from the silica during processing to produce silicon suitable for semiconductor applications.

In a silicon crystal, each atom shares electrons with its neighbors. These shared electrons form covalent bonds that hold the crystal lattice together. This bonding structure is the basis for understanding how controlled impurities, or dopants, can change the electrical properties of the material.

To create P-type silicon, silicon can be doped with a trivalent element such as boron. Because boron has three valence electrons rather than silicon's four, the substitution creates an electron deficiency. This deficiency is treated as a mobile hole, and the resulting semiconductor is called P-type material.

To create N-type silicon, a pentavalent element such as phosphorus can be introduced. Phosphorus has five valence electrons, leaving an additional electron that can participate in electrical conduction. Semiconductor material with an excess of electrons is called N-type material.

When P-type and N-type materials are brought together, electrons from the N-type side diffuse toward the P-type side and combine with holes. This charge redistribution produces an electrically charged region near the interface.

The region near the junction becomes depleted of mobile charge carriers and is therefore called the depletion region. The resulting electric field creates a potential barrier that opposes further diffusion of charge carriers. Equilibrium is reached when the opposing effects balance.

Biasing the PN Junction

The behavior of the PN junction can be controlled by applying an external voltage. The two basic conditions are reverse bias and forward bias.

Reverse bias

With reverse bias, the positive terminal of the DC source is connected to the N-type material and the negative terminal to the P-type material. The applied electric field drives the charge carriers away from the junction, increasing the depletion region and making carrier crossing more difficult. Under normal conditions, very little current flows.

Forward bias

With forward bias, the positive terminal is connected to the P-type material and the negative terminal to the N-type material. The applied voltage reduces the effective barrier at the junction, allowing electrons and holes to move toward the junction and recombine.

This ability to preferentially conduct current in one direction is one of the defining properties of a semiconductor diode. For optical devices, however, the most interesting consequence of forward bias is what happens when electrons and holes recombine in a suitable semiconductor material.

Compound Semiconductors and Light Emission

When a suitable compound semiconductor is used, electron-hole recombination can result in the emission of light. Important compound semiconductor families include combinations of Group 13 and Group 15 elements, such as gallium, aluminum, indium, phosphorus, arsenic, and related materials.

For example, gallium arsenide (GaAs) can be doped to create P-type and N-type regions. When carriers recombine in the appropriate device structure, energy can be released as photons rather than primarily as lattice vibrations.

A convenient way to describe this process is through the energy levels of electrons. When energy is supplied to an electron, it can move from the valence band to the higher-energy conduction band. When the electron returns to a lower energy state, the energy difference can be released as a photon.

Here, ΔE is the energy difference, h is Planck's constant, ν is the optical frequency, c is the speed of light in vacuum, and λ is the wavelength. The relationship shows why the semiconductor's energy-band structure is directly related to the wavelength of the emitted light.

Bandgap and Optical Wavelength

The relationship between bandgap energy and optical wavelength can be illustrated using GaAs. A GaAs bandgap of approximately 1.42 eV corresponds to an energy of about 2.28 × 10⁻¹⁹ J.

Using Planck's constant and the speed of light gives:

Solving for wavelength gives approximately:

This is close to the 850 nm wavelength region used for many short-reach optical communication applications. Device composition and fabrication parameters can be adjusted to produce emission at desired wavelengths. The key concept is that the semiconductor material and its bandgap determine the energy—and therefore the wavelength—of the emitted light.

Semiconductor material	Representative communication wavelength	Typical context
GaAs-based materials	Around 850 nm	Short-reach optical communication
InGaAsP-based materials	Around 1300 nm	Short-reach systems where chromatic dispersion is an important consideration
InGaAsP-based laser structures	Including the 1530–1570 nm C-band region	Optical networking and long-reach systems

The 850 nm region remains important for short-reach applications such as premises and data-center networks. InGaAsP-based materials are also important for longer-wavelength devices, including laser diodes used in optical networking.

Important Properties of Communication LEDs

Non-collimated light

LED output is generally not collimated. The emitted light travels over a range of directions, causing the beam to diverge relatively quickly. This makes it more difficult to efficiently couple LED output into a small single-mode fiber core. As a result, LEDs have traditionally been associated with multimode fiber, whose larger core makes coupling easier.

Incoherent light

LED light is also incoherent. In coherent light, the phase relationship between waves remains constant over time. In incoherent light, that phase relationship varies.

Wide spectral width

LEDs emit over a relatively broad range of wavelengths rather than concentrating their optical power into an extremely narrow spectral line. This wider spectral width makes the signal more susceptible to wavelength-dependent effects such as chromatic dispersion. Semiconductor lasers generally provide much narrower spectral widths.

Limited modulation speed

LEDs cannot normally be modulated as rapidly as semiconductor lasers. This limits their usefulness for higher-speed communication systems and has contributed to their replacement by laser-based transmitters in many applications.

Low cost

Cost has historically been an important advantage of LEDs. In short-reach systems, the lower cost of LED sources helped make multimode-fiber networks practical, particularly where large numbers of transmitter and receiver pairs were required.

LED characteristic	Impact on optical communications
Non-collimated output	Makes efficient coupling into small single-mode cores more difficult.
Incoherent emission	Not suitable for systems that depend on a stable optical phase relationship.
Broad spectral width	Increases sensitivity to chromatic-dispersion effects.
Lower modulation capability	Limits achievable transmission rates compared with laser-based transmitters.
Low device cost	Historically attractive for short-reach, multimode-fiber applications.

Summary

LEDs are used extensively in lighting and display technologies, but their role in optical communications is now much more limited. Their historical importance in short-reach fiber systems makes them useful for understanding the development of optical communications, while their underlying semiconductor physics provides a foundation for understanding laser diodes and other photonic devices.

The key concepts begin with the PN junction. Doping creates P-type and N-type regions; joining those regions produces a depletion region and a potential barrier; forward bias allows electrons and holes to approach and recombine; and, in suitable compound semiconductors, that recombination can produce photons. The semiconductor bandgap determines the energy of the emitted photons and therefore the wavelength of the light.