Light Generation and Sources
Light generation technologies form the foundation of optoelectronics, converting electrical energy into optical radiation across the spectrum from ultraviolet through visible to infrared wavelengths. These devices enable applications ranging from everyday illumination to precision manufacturing, communications, sensing, and scientific research.
Modern light sources have evolved from incandescent filaments and gas-discharge tubes to compact semiconductor devices that offer fine control over wavelength, intensity, beam quality, and modulation speed. Selecting the right source means weighing several characteristics at once: emission wavelength and spectral width, coherence, output power, electrical and optical efficiency, beam shape, modulation bandwidth, and lifetime. This category groups the subject into five device families, summarized below, and then develops the physics, figures of merit, drive requirements, and applications that they share.
Articles in This Category
Mechanisms of Light Generation
Three mechanisms account for nearly every practical source. Incandescence radiates a broad thermal spectrum from a hot body. Gas discharge excites atoms or molecules in a plasma, which then radiate on characteristic lines. Electroluminescence converts electrical carriers directly into photons in a semiconductor or an organic film. The first two dominated lighting for a century and survive where their particular spectra matter; the third underlies almost all of modern optoelectronics.
Spontaneous Emission and Electroluminescence
Light generation in semiconductor devices occurs through the radiative recombination of electrons and holes across a forward-biased junction. In a light-emitting diode this produces spontaneous emission: photons are emitted at random phases and directions, giving incoherent light with a relatively broad spectral width, typically on the order of 20 to 40 nanometers for a visible LED. Because the emission is isotropic inside a high-index crystal, much of it is trapped by total internal reflection, and extracting it requires shaped dies, roughened surfaces, patterned substrates, or encapsulants that raise the escape cone.
Stimulated Emission and the Laser Threshold
A laser adds an optical cavity that feeds emitted photons back through the gain region. An arriving photon can stimulate a second photon identical in phase, direction, polarization, and wavelength. Below a threshold current the cavity loses more light than the gain replaces and the device behaves as a weak, broadband emitter. Above threshold, stimulated emission dominates, the optical power climbs steeply and almost linearly with current, and the output becomes coherent and narrowband with a well-defined beam. The same physics underlies both device families; the presence and quality of the resonator is what separates a lamp-like emitter from a laser. The threshold itself is a useful diagnostic, since it rises with temperature and with accumulated degradation.
Loss Mechanisms
The efficiency of the conversion depends on material quality, device geometry, and operating conditions. Non-radiative recombination at defects and surfaces, Auger recombination, carrier leakage over the confining barriers, series resistance, optical absorption, and poor extraction all divert energy away from useful emission. In LEDs these mechanisms combine to produce efficiency droop, a decline in efficiency as drive current density rises, which is why high-power luminaires generally use many emitters at moderate current rather than a few driven hard.
Semiconductor Materials and Wavelength
The wavelength of emitted light is set primarily by the bandgap energy of the semiconductor, since each recombination event releases a photon of roughly that energy. The relationship is convenient in practice: wavelength in nanometers is approximately 1240 divided by the bandgap in electron volts, so a 2.5 eV gap emits near 500 nm. Engineers tune the bandgap by choosing the compound and adjusting its alloy composition, then confine carriers and photons with heterostructures and quantum wells so that recombination happens where it is wanted.
Several material families cover the practical spectrum:
- Gallium arsenide (GaAs) and aluminum gallium arsenide (AlGaAs) emit in the red and near-infrared, roughly 650 to 900 nm. They supply the 850 nm VCSELs used in short-reach data links and the 808 nm pump diodes used for solid-state lasers.
- Indium gallium arsenide phosphide (InGaAsP) and indium gallium arsenide (InGaAs) on indium phosphide reach the 1310 nm and 1550 nm telecommunications windows, where silica fiber has its dispersion minimum and its loss minimum respectively.
- Aluminum indium gallium phosphide (AlInGaP) covers red, orange, and yellow with high efficiency, and loses efficiency toward the green as its confinement barriers become too shallow to hold carriers.
- Gallium nitride (GaN) and indium gallium nitride (InGaN) opened up efficient blue and violet emission, enabling phosphor-converted white LEDs and blue-violet laser diodes. That work was recognized by the 2014 Nobel Prize in Physics, awarded to Isamu Akasaki, Hiroshi Amano, and Shuji Nakamura for the invention of efficient blue light-emitting diodes.
- Aluminum gallium nitride (AlGaN) extends emission into the ultraviolet, including the 265 to 280 nm band used for disinfection. External quantum efficiency there remains low, commonly only a few percent, because of poor p-type doping, strong optical absorption, and difficult light extraction.
One gap in this coverage is important enough to have its own name. Both the nitride and phosphide families lose efficiency in the green and yellow-green region, roughly 500 to 570 nm, a shortfall known as the green gap. It arises in the nitrides from the strong internal electric fields and indium composition fluctuations that accompany high indium content, and in the phosphides from carrier leakage. The practical consequence is that direct green emitters trail blue and red devices in efficiency, which is one reason white LEDs are built from a blue emitter plus a phosphor rather than from mixed red, green, and blue chips. Achieving any of these results depends on high-quality epitaxial growth, careful doping, and lattice-matched or strain-managed layer structures.
Optical Cavity and Resonator Design
Laser sources require an optical cavity that provides the feedback needed to sustain oscillation. Edge-emitting laser diodes use cleaved crystal facets as partially reflecting mirrors, with light emerging from the side of the chip in an elliptical, strongly diverging beam that usually needs a fast-axis collimating lens. Vertical-cavity surface-emitting lasers (VCSELs) instead stack distributed Bragg reflectors above and below a very short vertical cavity, producing a low-divergence circular beam, low threshold current, and a device that can be tested on-wafer and arrayed in two dimensions. Distributed-feedback (DFB) lasers embed a grating along the cavity to enforce single-longitudinal-mode operation, giving the narrow linewidth that long-haul fiber links require. Solid-state and gas lasers use discrete external mirrors in linear, folded, or ring configurations, and fiber lasers replace those mirrors with fiber Bragg gratings written into the waveguide itself.
In every case the cavity geometry governs threshold, output power, beam divergence, transverse and longitudinal mode structure, spectral purity, and noise. Cavity length sets the longitudinal mode spacing, so a short VCSEL cavity naturally supports a single wavelength while a long edge emitter needs a grating to select one. The mirror reflectivities trade output coupling against threshold. Optical feedback from a downstream reflection can destabilize a diode laser badly enough that precision sources are supplied with an integrated optical isolator. The optical materials and components used for these mirrors, coatings, and windows must also survive the intracavity intensity, which is often far higher than the emitted beam.
Figures of Merit
A consistent set of parameters describes source performance and allows fair comparison across technologies.
- Center wavelength and spectral width. A laser diode may emit a line narrower than a picometer; an LED emits tens of nanometers; a superluminescent diode or supercontinuum source deliberately emits far more. Wavelength also drifts with temperature and drive current, which matters wherever a filter, grating, or fluorophore must be matched.
- Coherence. Temporal coherence, expressed as coherence length, determines whether a source can support interferometry or holography. Spatial coherence determines how tightly the beam can be focused. Both are assets in metrology and liabilities in imaging, where coherent illumination produces speckle.
- Radiant flux and luminous flux. Radiant flux in watts measures total optical power; luminous flux in lumens weights that power by the human eye's photopic response, which peaks at 555 nm where 1 watt of radiation corresponds to 683 lumens. Infrared and ultraviolet sources are therefore rated in watts, and lighting products in lumens.
- Wall-plug efficiency. The ratio of optical output power to electrical input power, sometimes separated into internal quantum efficiency, extraction efficiency, and electrical efficiency. Lighting products often quote luminous efficacy in lumens per watt instead, which folds in the eye response.
- Beam quality. The beam parameter product, or the equivalent M-squared factor, states how closely a beam approaches the diffraction limit. A single-mode fiber laser near M-squared of 1 can be focused to a spot a few tens of micrometers across, which is what makes precision cutting possible; a broad-area diode bar cannot. For incoherent sources the analogous constraint is étendue, the product of source area and emission solid angle, which optics can redistribute but never reduce.
- Modulation bandwidth. LEDs are limited by carrier lifetime to roughly tens of megahertz, and phosphor-converted white LEDs to far less, because the phosphor's decay time smears fast transitions. That is adequate for remote controls and visible-light communication. Directly modulated laser diodes reach tens of gigahertz, and higher rates use an external modulator to avoid the wavelength chirp that direct modulation causes.
- Noise and stability. Relative intensity noise, linewidth, mode hopping, and pointing stability set the floor for interferometric and spectroscopic measurements, and for analog optical links.
- Lifetime. LEDs degrade gradually rather than failing abruptly, so life is quoted as lumen maintenance: L70 is the operating time at which output has fallen to 70 percent of its initial value. The IES LM-80 method measures lumen maintenance under controlled temperature and current, and IES TM-21 projects those measurements forward, capping any projection at six times the test duration so that a 10,000-hour test supports no more than a 60,000-hour claim. Laser diodes are instead characterized by mean time to failure at a stated power and temperature, with catastrophic optical damage at the facet as an additional abrupt failure mode.
Driving and Controlling Light Sources
Emitters are current-driven devices, and the drive circuit is as much a part of the design as the emitter. An LED's forward voltage falls as it warms, so a source that holds voltage constant will deliver more current into a hotter device, which heats it further; constant-current drive removes that positive feedback. Dimming is done either by pulse-width modulation, which holds chromaticity nearly constant because the diode always operates at the same current, or by analog current reduction, which is flicker-free but shifts color slightly. Switching drivers must also be designed with flicker and electromagnetic compatibility in mind, since the modulation is visible to cameras and to some observers. Driver topologies and dimming methods are developed under light-emitting diodes (LEDs).
Laser diodes demand more care. Above threshold the slope of power against current is steep, so a small current overshoot can destroy the facet; drivers therefore include soft-start, current limiting, and protection against transients at power-up and power-down. Most packages integrate a monitor photodiode so that an automatic power control loop can hold optical output constant as temperature and aging change the threshold. Wavelength-critical sources add a thermoelectric cooler and a thermistor to stabilize the junction, because a DFB laser tunes on the order of 0.1 nanometer per kelvin. Laser diodes are also highly sensitive to electrostatic discharge, which makes grounded handling and input protection routine rather than optional. Driver architectures and control loops for these sources are covered under laser diodes and semiconductor lasers.
Efficiency and Thermal Management
Every watt that is not emitted appears as heat at the junction, and junction temperature feeds back into nearly every other parameter. As a diode heats, its output falls, its wavelength drifts, its threshold rises, and its degradation accelerates. The best commercial white LED packages are specified above 200 lumens per watt at moderate drive currents, and a laboratory device reached 303 lumens per watt at 350 milliamperes and a correlated color temperature near 5150 K in 2014, against a theoretical ceiling of 683 lumens per watt for monochromatic green light and a considerably lower practical ceiling for broadband white. High-power 9xx nm pump diodes, the most electrically efficient lasers made, convert roughly 60 to 70 percent of their input power to light in commercial products. Fiber lasers built from those pumps typically reach 30 to 40 percent wall-plug efficiency at the system level, and carbon-dioxide lasers considerably less.
Effective thermal design is therefore essential rather than incidental. Metal-core printed circuit boards, ceramic submounts, heat sinks, heat spreaders, and forced air or liquid cooling all serve to lower the thermal resistance from junction to ambient, and wavelength-critical sources add thermoelectric coolers for active stabilization. Because manufacturers specify performance and lifetime at a given junction temperature, the thermal path is what determines whether a datasheet number is achievable in the finished product. The broader techniques are covered under thermal management.
Illumination and Display
LED lighting has become the dominant technology for general illumination, combining high efficacy, long rated lifetimes, dimmability, instant start, and design flexibility. Most white LEDs pair a blue InGaN emitter with a phosphor, commonly cerium-doped yttrium aluminum garnet, that down-converts part of the blue light to longer wavelengths; the phosphor mix sets the correlated color temperature and the color-rendering index, and the unavoidable Stokes loss in that conversion is one reason white efficacy trails blue. Automotive headlamps, architectural and street lighting, and horticultural fixtures tuned to chlorophyll absorption all exploit these characteristics, and the field is treated in depth under lighting systems and applications.
In displays, LEDs serve both as edge and direct backlights for liquid-crystal panels and, increasingly, as the image-forming elements themselves. Mini-LED backlights divide the panel into hundreds or thousands of independently dimmed zones to raise contrast, while micro-LED arrays place an individual emitter at each subpixel. OLED technology, which emits directly from organic layers, enables thin, flexible, high-contrast panels and diffuse area lighting. Quantum-dot films add a narrowband conversion layer that widens the color gamut. These architectures are compared under display technologies.
Communications
Laser diodes are the workhorses of optical communication, modulated at multi-gigabit-per-second rates to carry the bulk of global data traffic over fiber. DFB and externally modulated lasers serve long-haul and metro links at 1550 nm, where narrow linewidth limits dispersion penalties and erbium-doped fiber amplifiers can boost the signal without converting it back to electronics. Arrays of 850 nm VCSELs provide low-cost, low-power, high-density interconnects inside data centers, where reach is short enough that multimode fiber suffices. Wavelength-division multiplexing combines many sources on a single fiber, multiplying capacity without adding fiber. Free-space optical links use laser sources for high-bandwidth wireless transmission between buildings or satellites, and LED-based visible-light communication offers a complement for indoor positioning and short-range data. The receivers that complete these links are covered under optical communication systems.
Industrial and Manufacturing
High-power lasers are indispensable tools for cutting, welding, drilling, marking, and additive manufacturing. Fiber lasers, with their excellent beam quality, high electrical efficiency, and fiber-delivered output, have become the leading choice for precision metal cutting and welding at kilowatt power levels. Carbon-dioxide lasers emitting near 10.6 micrometers remain widely used for non-metals such as plastics, wood, acrylic, and textiles, which absorb strongly in the far infrared. Blue and green diode lasers have found a niche in welding copper and other highly reflective metals, which absorb short wavelengths far better than the near infrared and so waste less of the incident power.
In semiconductor fabrication, deep-ultraviolet lithography uses argon-fluoride excimer lasers at 193 nm to pattern circuits, while the most advanced nodes rely on extreme-ultraviolet systems that generate 13.5 nm light by striking tin droplets with a high-power carbon-dioxide laser to form a plasma. LED and laser illuminators also provide the bright, stable, spectrally controlled light that machine-vision and inspection systems need for consistent imaging. These processes are developed further under industrial and scientific applications, and the related sensing systems under LIDAR and active optical sensing, where pulsed edge emitters and VCSEL arrays supply the illumination.
Medical and Scientific
Medical applications span therapeutic lasers in surgery, ophthalmology, and dermatology, where wavelength is chosen so that the target chromophore, whether water, hemoglobin, or melanin, absorbs strongly while surrounding tissue does not. Diagnostic and laboratory instruments use laser sources for flow cytometry, sequencing, and spectroscopy, and broadband or superluminescent sources feed optical coherence tomography, whose depth resolution improves as source bandwidth increases. Research frequently demands tightly specified wavelength, bandwidth, pulse duration, or peak power: ultrafast mode-locked lasers reach femtosecond pulses for nonlinear imaging, micromachining, and precision metrology, while tunable lasers and optical frequency combs support spectroscopy and optical timekeeping. The clinical side of this work is developed under biomedical photonics.
Safety and Standards
Sources bright enough to be useful are often bright enough to injure, and the eye is the organ at risk because it focuses collimated light onto the retina. IEC 60825-1 classifies laser products by accessible emission across 180 nm to 1 mm. Class 1 is safe under all reasonably foreseeable conditions, including enclosed high-power systems whose beam is inaccessible in normal use. Class 2 covers visible beams that are safe because of the blink reflex. Class 3R marks a modest hazard, Class 3B a beam that is hazardous on direct viewing, and Class 4 a beam that can injure through diffuse reflection and can ignite materials; Classes 3B and 4 require interlocks, key control, warning signs, and eyewear matched to the wavelength. An M suffix, as in Class 1M or 2M, marks a product that is safe as emitted but hazardous if viewed through magnifying optics. Invisible near-infrared beams deserve particular caution, since the blink reflex offers no protection at wavelengths the eye cannot see.
Non-laser sources fall under IEC 62471, which sorts lamps and LED products into photobiological risk groups by ultraviolet, blue-light, and infrared hazard. Ultraviolet-C disinfection fixtures and high-power blue emitters both merit shielding and access control. Beyond safety, performance standards govern how sources are measured and specified, including LM-80 and TM-21 for LED lumen maintenance, so that competing products can be compared on the same basis.
Selecting a Source
Source choice follows from a short sequence of questions, and answering them in order usually narrows the field to one or two candidates.
- What wavelength, and how pure? The required band eliminates most materials at once, and the tolerance on center wavelength decides whether a free-running emitter will do or a stabilized DFB, external-cavity, or gas laser is needed.
- Coherent or incoherent? Interferometry, holography, and long-distance projection need coherence; illumination and imaging usually suffer from it, because speckle degrades the image.
- How much power, and in what form? Continuous output, pulse energy, and peak power are different requirements, and a source optimized for one is rarely good at another.
- How tightly must it focus? Beam quality and étendue decide whether the power can be delivered to a small spot. No optical system recovers what a poor beam has already lost.
- How fast must it switch? Steady illumination allows almost any source; gigabit data or nanosecond ranging forces a laser diode and a matched driver.
- What is the thermal and electrical environment? Available supply, ambient temperature, cooling budget, and vibration routinely rule out the technically ideal emitter in favor of a robust one.
- What lifetime and safety class are acceptable? A sealed Class 1 product costs far less to deploy than a Class 4 system that requires a controlled area, and lumen maintenance targets shape both the drive current and the heat sink.
Trends and Outlook
Light generation continues to advance, driven by demand for higher efficiency, new wavelengths, better beam quality, and novel applications. Solid-state lighting has already followed a decades-long trajectory of falling cost per lumen and rising output per package, an observation known as Haitz's law, and attention has shifted from raw efficacy toward spectral quality, glare, controllability, and the human response to light. Ultraviolet-C emitters are the current frontier in efficiency, where external quantum efficiency remains low enough that improvement would open large disinfection markets. Micro-LED displays face a manufacturing rather than a physics problem, namely the mass transfer and repair of millions of tiny emitters.
On the laser side, heterogeneous and hybrid integration is placing III-V gain material directly onto silicon photonic circuits, so that sources, modulators, and waveguides share a die. Quantum-dot lasers grown on silicon promise temperature-insensitive operation, and narrow-linewidth integrated lasers support coherent communication and precision sensing. Related developments appear under emerging photonic technologies and quantum photonics, where single-photon and entangled-photon sources form a device class of their own.
Conclusion
Every source in this category performs the same conversion of electrical energy into photons, yet the engineering diverges sharply once wavelength, coherence, power, beam quality, and switching speed are fixed. An LED answers most illumination and indication problems; a laser diode answers most communication, ranging, and precision-delivery problems; and solid-state, fiber, gas, specialty, and comb sources answer the problems that need power, pulse energy, bandwidth, or absolute frequency accuracy that a diode cannot supply. Understanding bandgap engineering, cavity design, efficiency limits, and thermal behavior is what makes the choice between them systematic rather than arbitrary. The five subcategories above develop each device family in detail, and the companion category on light detection and sensing covers the detectors that these emitters are almost always paired with.