From Wikipedia, the free encyclopedia

“LED” redirects here. For other uses, see LED (disambiguation).This article is about the basics of light emitting diodes. For application to area lighting, see LED lamp.Not to be confused with LCD.

A light-emitting diode (LED) is a semiconductor light source that emits light when current flows through it. Electrons in the semiconductor recombine with electron holes, releasing energy in the form of photons. The color of the light (corresponding to the energy of the photons) is determined by the energy required for electrons to cross the band gap of the semiconductor.^[5] White light is obtained by using multiple semiconductors or a layer of light-emitting phosphor on the semiconductor device.^[6]

Appearing as practical electronic components in 1962, the earliest LEDs emitted low-intensity infrared light.^[7] Infrared LEDs are used in remote-control circuits, such as those used with a wide variety of consumer electronics. The first visible-light LEDs were of low intensity and limited to red. Modern LEDs are available across the visible, ultraviolet, and infrared wavelengths, with high light output.

Early LEDs were often used as indicator lamps, replacing small incandescent bulbs, and in seven-segment displays. Recent developments have produced high-output white light LEDs suitable for room and outdoor area lighting. LEDs have led to new displays and sensors, while their high switching rates are useful in advanced communications technology.

LEDs have many advantages over incandescent light sources, including lower energy consumption, longer lifetime, improved physical robustness, smaller size, and faster switching. LEDs are used in applications as diverse as aviation lighting, automotive headlamps, advertising, general lighting, traffic signals, camera flashes, lighted wallpaper, plant growing light, and medical devices.^[8]

Unlike a laser, the light emitted from an LED is neither spectrally coherent nor even highly monochromatic. However its spectrum is sufficiently narrow that it appears to the human eye as a pure (saturated) color.^[9][10] Nor, unlike most lasers, is its radiation spatially coherent, so that it cannot approach the very high brightnesses characteristic of lasers.

History

Discoveries and early devices

Electroluminescence as a phenomenon was discovered in 1907 by the British experimenter H. J. Round of Marconi Labs, using a crystal of silicon carbide and a cat’s-whisker detector.^[11][12] Russian inventor Oleg Losev reported creation of the first LED in 1927.^[13] His research was distributed in Soviet, German and British scientific journals, but no practical use was made of the discovery for several decades.^[14][15]

In 1936, Georges Destriau observed that electroluminescence could be produced when zinc sulphide (ZnS) powder is suspended in an insulator and an alternating electrical field is applied to it. In his publications, Destriau often referred to luminescence as Losev-Light. Destriau worked in the laboratories of Madame Marie Curie, also an early pioneer in the field of luminescence with research on radium.^[16][17]

Hungarian Zoltán Bay together with György Szigeti pre-empted LED lighting in Hungary in 1939 by patenting a lighting device based on SiC, with an option on boron carbide, that emitted white, yellowish white, or greenish white depending on impurities present.^[18]

Kurt Lehovec, Carl Accardo, and Edward Jamgochian explained these first LEDs in 1951 using an apparatus employing SiC crystals with a current source of a battery or a pulse generator and with a comparison to a variant, pure, crystal in 1953.^[19][20]

Rubin Braunstein^[21] of the Radio Corporation of America reported on infrared emission from gallium arsenide (GaAs) and other semiconductor alloys in 1955.^[22] Braunstein observed infrared emission generated by simple diode structures using gallium antimonide (GaSb), GaAs, indium phosphide (InP), and silicon-germanium (SiGe) alloys at room temperature and at 77 kelvins.

In 1957, Braunstein further demonstrated that the rudimentary devices could be used for non-radio communication across a short distance. As noted by Kroemer^[23] Braunstein “…had set up a simple optical communications link: Music emerging from a record player was used via suitable electronics to modulate the forward current of a GaAs diode. The emitted light was detected by a PbS diode some distance away. This signal was fed into an audio amplifier and played back by a loudspeaker. Intercepting the beam stopped the music. We had a great deal of fun playing with this setup.” This setup presaged the use of LEDs for optical communication applications.

In September 1961, while working at Texas Instruments in Dallas, Texas, James R. Biard and Gary Pittman discovered near-infrared (900 nm) light emission from a tunnel diode they had constructed on a GaAs substrate.^[7] By October 1961, they had demonstrated efficient light emission and signal coupling between a GaAs p-n junction light emitter and an electrically isolated semiconductor photodetector.^[24] On August 8, 1962, Biard and Pittman filed a patent titled “Semiconductor Radiant Diode” based on their findings, which described a zinc-diffused p–n junction LED with a spaced cathode contact to allow for efficient emission of infrared light under forward bias. After establishing the priority of their work based on engineering notebooks predating submissions from G.E. Labs, RCA Research Labs, IBM Research Labs, Bell Labs, and Lincoln Lab at MIT, the U.S. patent office issued the two inventors the patent for the GaAs infrared (IR) light-emitting diode (U.S. Patent US3293513), the first practical LED.^[7] Immediately after filing the patent, Texas Instruments (TI) began a project to manufacture infrared diodes. In October 1962, TI announced the first commercial LED product (the SNX-100), which employed a pure GaAs crystal to emit an 890 nm light output.^[7] In October 1963, TI announced the first commercial hemispherical LED, the SNX-110.^[25]

The first visible-spectrum (red) LED was developed in 1962 by Nick Holonyak, Jr. while working at General Electric. Holonyak first reported his LED in the journal Applied Physics Letters on December 1, 1962.^[26][27] M. George Craford,^[28] a former graduate student of Holonyak, invented the first yellow LED and improved the brightness of red and red-orange LEDs by a factor of ten in 1972.^[29] In 1976, T. P. Pearsall designed the first high-brightness, high-efficiency LEDs for optical fiber telecommunications by inventing new semiconductor materials specifically adapted to optical fiber transmission wavelengths.^[30]

Initial commercial development

The first commercial visible-wavelength LEDs were commonly used as replacements for incandescent and neon indicator lamps, and in seven-segment displays,^[31] first in expensive equipment such as laboratory and electronics test equipment, then later in such appliances as calculators, TVs, radios, telephones, as well as watches (see list of signal uses). Until 1968, visible and infrared LEDs were extremely costly, in the order of US$200 per unit, and so had little practical use.^[32]

Hewlett-Packard (HP) was engaged in research and development (R&D) on practical LEDs between 1962 and 1968, by a research team under Howard C. Borden, Gerald P. Pighini and Mohamed M. Atalla at HP Associates and HP Labs.^[33] During this time, Atalla launched a material science investigation program on gallium arsenide (GaAs), gallium arsenide phosphide (GaAsP) and indium arsenide (InAs) devices at HP,^[34] and they collaborated with Monsanto Company on developing the first usable LED products.^[35] The first usable LED products were HP’s LED display and Monsanto’s LED lamp, both launched in 1968.^[35] Monsanto was the first organization to mass-produce visible LEDs, using GaAsP in 1968 to produce red LEDs suitable for indicators.^[32] Monsanto had previously offered to supply HP with GaAsP, but HP decided to grow its own GaAsP.^[32] In February 1969, Hewlett-Packard introduced the HP Model 5082-7000 Numeric Indicator, the first LED device to use integrated circuit technology.^[33] It was the first intelligent LED display, and was a revolution in digital display technology, replacing the Nixie tube and becoming the basis for later LED displays.^[36]

Atalla left HP and joined Fairchild Semiconductor in 1969.^[37] He was the vice president and general manager of the Microwave & Optoelectronics division,^[38] from its inception in May 1969 up until November 1971.^[39] He continued his work on LEDs, proposing they could be used for indicator lights and optical readers in 1971.^[40] In the 1970s, commercially successful LED devices at less than five cents each were produced by Fairchild Optoelectronics. These devices employed compound semiconductor chips fabricated with the planar process (developed by Jean Hoerni,^[41][42] based on Atalla’s surface passivation method^[43][44]). The combination of planar processing for chip fabrication and innovative packaging methods enabled the team at Fairchild led by optoelectronics pioneer Thomas Brandt to achieve the needed cost reductions.^[45] LED producers continue to use these methods.^[46]

The early red LEDs were bright enough only for use as indicators, as the light output was not enough to illuminate an area. Readouts in calculators were so small that plastic lenses were built over each digit to make them legible. Later, other colors became widely available and appeared in appliances and equipment.

Early LEDs were packaged in metal cases similar to those of transistors, with a glass window or lens to let the light out. Modern indicator LEDs are packed in transparent molded plastic cases, tubular or rectangular in shape, and often tinted to match the device color. Infrared devices may be dyed, to block visible light. More complex packages have been adapted for efficient heat dissipation in high-power LEDs. Surface-mounted LEDs further reduce the package size. LEDs intended for use with fiber optics cables may be provided with an optical connector.

Blue LED

The first blue-violet LED using magnesium-doped gallium nitride was made at Stanford University in 1972 by Herb Maruska and Wally Rhines, doctoral students in materials science and engineering.^[47][48] At the time Maruska was on leave from RCA Laboratories, where he collaborated with Jacques Pankove on related work. In 1971, the year after Maruska left for Stanford, his RCA colleagues Pankove and Ed Miller demonstrated the first blue electroluminescence from zinc-doped gallium nitride, though the subsequent device Pankove and Miller built, the first actual gallium nitride light-emitting diode, emitted green light.^[49][50] In 1974 the U.S. Patent Office awarded Maruska, Rhines and Stanford professor David Stevenson a patent for their work in 1972 (U.S. Patent US3819974 A). Today, magnesium-doping of gallium nitride remains the basis for all commercial blue LEDs and laser diodes. In the early 1970s, these devices were too dim for practical use, and research into gallium nitride devices slowed.

In August 1989, Cree introduced the first commercially available blue LED based on the indirect bandgap semiconductor, silicon carbide (SiC).^[51] SiC LEDs had very low efficiency, no more than about 0.03%, but did emit in the blue portion of the visible light spectrum.^[52][53]

In the late 1980s, key breakthroughs in GaN epitaxial growth and p-type doping^[54] ushered in the modern era of GaN-based optoelectronic devices. Building upon this foundation, Theodore Moustakas at Boston University patented a method for producing high-brightness blue LEDs using a new two-step process in 1991.^[55]

Two years later, in 1993, high-brightness blue LEDs were demonstrated by Shuji Nakamura of Nichia Corporation using a gallium nitride growth process.^[56][57][58] In parallel, Isamu Akasaki and Hiroshi Amano in Nagoya were working on developing the important GaN deposition on sapphire substrates and the demonstration of p-type doping of GaN. This new development revolutionized LED lighting, making high-power blue light sources practical, leading to the development of technologies like Blu-ray.

Nakamura was awarded the 2006 Millennium Technology Prize for his invention.^[59] Nakamura, Hiroshi Amano and Isamu Akasaki were awarded the Nobel Prize in Physics in 2014 for the invention of the blue LED.^[60] In 2015, a US court ruled that three companies had infringed Moustakas’s prior patent, and ordered them to pay licensing fees of not less than US$13 million.^[61]

In 1995, Alberto Barbieri at the Cardiff University Laboratory (GB) investigated the efficiency and reliability of high-brightness LEDs and demonstrated a “transparent contact” LED using indium tin oxide (ITO) on (AlGaInP/GaAs).

In 2001^[62] and 2002,^[63] processes for growing gallium nitride (GaN) LEDs on silicon were successfully demonstrated. In January 2012, Osram demonstrated high-power InGaN LEDs grown on silicon substrates commercially,^[64] and GaN-on-silicon LEDs are in production at Plessey Semiconductors. As of 2017, some manufacturers are using SiC as the substrate for LED production, but sapphire is more common, as it has the most similar properties to that of gallium nitride, reducing the need for patterning the sapphire wafer (patterned wafers are known as epi wafers). Samsung, the University of Cambridge, and Toshiba are performing research into GaN on Si LEDs. Toshiba has stopped research, possibly due to low yields.^{[65][66][67][68][69][70][71]} Some opt towards epitaxy, which is difficult on silicon, while others, like the University of Cambridge, opt towards a multi-layer structure, in order to reduce (crystal) lattice mismatch and different thermal expansion ratios, in order to avoid cracking of the LED chip at high temperatures (e.g. during manufacturing), reduce heat generation and increase luminous efficiency. Epitaxy (or patterned sapphire) can be carried out with nanoimprint lithography.^{[72][73][74][75][76][77][78]} GaN is often deposited using Metalorganic vapour-phase epitaxy (MOCVD).

White LEDs and the illumination breakthrough

Even though white light can be created using individual red, green and blue LEDs, this results in poor color rendering, since only three narrow bands of wavelengths of light are being emitted. The attainment of high efficiency blue LEDs was quickly followed by the development of the first white LED. In this device a Y
₃Al
₅O
₁₂:Ce (known as “YAG” or Ce:YAG phosphor) cerium doped phosphor coating produces yellow light through fluorescence. The combination of that yellow with remaining blue light appears white to the eye. Using different phosphors produces green and red light through fluorescence. The resulting mixture of red, green and blue is perceived as white light, with improved color rendering compared to wavelengths from the blue LED/YAG phosphor combination.

The first white LEDs were expensive and inefficient. However, the light output of LEDs has increased exponentially. The latest research and development has been propagated by Japanese manufacturers such as Panasonic, and Nichia, and by Korean and Chinese manufacturers such as Samsung, Kingsun, and others. This trend in increased output has been called Haitz’s law after Dr. Roland Haitz.^[79]

Light output and efficiency of blue and near-ultraviolet LEDs rose and the cost of reliable devices fell. This led to relatively high-power white-light LEDs for illumination, which are replacing incandescent and fluorescent lighting.^[80][81]

Experimental white LEDs have been demonstrated to produce 303 lumens per watt of electricity (lm/w); some can last up to 100,000 hours.^[82][83] However, commercially available LEDs have an efficiency of up to 223 lm/w.^[84][85][86] Compared to incandescent bulbs, this is not only a huge increase in electrical efficiency, and even though the bulbs are more expensive to purchase, significantly cheaper overall cost per bulb.^[87]

The LED chip is encapsulated inside a small, plastic, white mold. It can be encapsulated using resin, silicone, or epoxy containing (powdered) Cerium doped YAG phosphor. After allowing the solvents to evaporate, the LEDs are often tested, and placed on tapes for SMT placement equipment for use in LED light bulb production. Encapsulation is performed after probing, dicing, die transfer from wafer to package, and wire bonding or flip chip mounting, perhaps using Indium tin oxide, a transparent electrical conductor. In this case, the bond wire(s) are attached to the ITO film that has been deposited in the LEDs. Some “remote phosphor” LED light bulbs use a single plastic cover with YAG phosphor for several blue LEDs, instead of using phosphor coatings on single chip white LEDs.

Colors

By selection of different semiconductor materials, single-color LEDs can be made that emit light in a narrow band of wavelengths from near-infrared through the visible spectrum and into the ultraviolet range. As the wavelengths become shorter, because of the larger band gap of these semiconductors, the operating voltage of the LED increases.

Blue and ultraviolet

Blue LEDs have an active region consisting of one or more InGaN quantum wells sandwiched between thicker layers of GaN, called cladding layers. By varying the relative In/Ga fraction in the InGaN quantum wells, the light emission can in theory be varied from violet to amber.

Aluminium gallium nitride (AlGaN) of varying Al/Ga fraction can be used to manufacture the cladding and quantum well layers for ultraviolet LEDs, but these devices have not yet reached the level of efficiency and technological maturity of InGaN/GaN blue/green devices. If un-alloyed GaN is used in this case to form the active quantum well layers, the device emits near-ultraviolet light with a peak wavelength centred around 365 nm. Green LEDs manufactured from the InGaN/GaN system are far more efficient and brighter than green LEDs produced with non-nitride material systems, but practical devices still exhibit efficiency too low for high-brightness applications.^{[citation needed]}

With AlGaN and AlGaInN, even shorter wavelengths are achievable. Near-UV emitters at wavelengths around 360–395 nm are already cheap and often encountered, for example, as black light lamp replacements for inspection of anti-counterfeiting UV watermarks in documents and bank notes, and for UV curing. While substantially more expensive, shorter-wavelength diodes are commercially available for wavelengths down to 240 nm.^[88] As the photosensitivity of microorganisms approximately matches the absorption spectrum of DNA, with a peak at about 260 nm, UV LED emitting at 250–270 nm are expected in prospective disinfection and sterilization devices. Recent research has shown that commercially available UVA LEDs (365 nm) are already effective disinfection and sterilization devices.^[89] UV-C wavelengths were obtained in laboratories using aluminium nitride (210 nm),^[90] boron nitride (215 nm)^[91][92] and diamond (235 nm).^[93]

White

There are two primary ways of producing white light-emitting diodes. One is to use individual LEDs that emit three primary colors—red, green and blue—and then mix all the colors to form white light. The other is to use a phosphor material to convert monochromatic light from a blue or UV LED to broad-spectrum white light, similar to a fluorescent lamp. The yellow phosphor is cerium-doped YAG crystals suspended in the package or coated on the LED. This YAG phosphor causes white LEDs to look yellow when off.

The ‘whiteness’ of the light produced is engineered to suit the human eye. Because of metamerism, it is possible to have quite different spectra that appear white. However, the appearance of objects illuminated by that light may vary as the spectrum varies. This is the issue of color rendition, quite separate from color temperature. An orange or cyan object could appear with the wrong color and much darker as the LED or phosphor does not emit the wavelength it reflects. The best color rendition LEDs use a mix of phosphors, resulting in less efficiency but better color rendering.

RGB systems

Mixing red, green, and blue sources to produce white light needs electronic circuits to control the blending of the colors. Since LEDs have slightly different emission patterns, the color balance may change depending on the angle of view, even if the RGB sources are in a single package, so RGB diodes are seldom used to produce white lighting. Nonetheless, this method has many applications because of the flexibility of mixing different colors,^[94] and in principle, this mechanism also has higher quantum efficiency in producing white light.^[95]

There are several types of multicolor white LEDs: di-, tri-, and tetrachromatic white LEDs. Several key factors that play among these different methods include color stability, color rendering capability, and luminous efficacy. Often, higher efficiency means lower color rendering, presenting a trade-off between the luminous efficacy and color rendering. For example, the dichromatic white LEDs have the best luminous efficacy (120 lm/W), but the lowest color rendering capability. However, although tetrachromatic white LEDs have excellent color rendering capability, they often have poor luminous efficacy. Trichromatic white LEDs are in between, having both good luminous efficacy (>70 lm/W) and fair color rendering capability.

One of the challenges is the development of more efficient green LEDs. The theoretical maximum for green LEDs is 683 lumens per watt but as of 2010 few green LEDs exceed even 100 lumens per watt. The blue and red LEDs approach their theoretical limits.

Multicolor LEDs also offer a new means to form light of different colors. Most perceivable colors can be formed by mixing different amounts of three primary colors. This allows precise dynamic color control. However, this type of LED’s emission power decays exponentially with rising temperature,^[96] resulting in a substantial change in color stability. Such problems inhibit industrial use. Multicolor LEDs without phosphors cannot provide good color rendering because each LED is a narrowband source. LEDs without phosphor, while a poorer solution for general lighting, are the best solution for displays, either backlight of LCD, or direct LED based pixels.

Dimming a multicolor LED source to match the characteristics of incandescent lamps is difficult because manufacturing variations, age, and temperature change the actual color value output. To emulate the appearance of dimming incandescent lamps may require a feedback system with color sensor to actively monitor and control the color.^[97]

Advantages

• Efficiency: LEDs emit more lumens per watt than incandescent light bulbs.^[127] The efficiency of LED lighting fixtures is not affected by shape and size, unlike fluorescent light bulbs or tubes.
• Color: LEDs can emit light of an intended color without using any color filters as traditional lighting methods need. This is more efficient and can lower initial costs.
• Size: LEDs can be very small (smaller than 2 mm^2[128]) and are easily attached to printed circuit boards.
• Warmup time: LEDs light up very quickly. A typical red indicator LED achieves full brightness in under a microsecond.^[129] LEDs used in communications devices can have even faster response times.
• Cycling: LEDs are ideal for uses subject to frequent on-off cycling, unlike incandescent and fluorescent lamps that fail faster when cycled often, or high-intensity discharge lamps (HID lamps) that require a long time before restarting.
• Dimming: LEDs can very easily be dimmed either by pulse-width modulation or lowering the forward current.^[130] This pulse-width modulation is why LED lights, particularly headlights on cars, when viewed on camera or by some people, seem to flash or flicker. This is a type of stroboscopic effect.
• Cool light: In contrast to most light sources, LEDs radiate very little heat in the form of IR that can cause damage to sensitive objects or fabrics. Wasted energy is dispersed as heat through the base of the LED.
• Slow failure: LEDs mainly fail by dimming over time, rather than the abrupt failure of incandescent bulbs.^[131]
• Lifetime: LEDs can have a relatively long useful life. One report estimates 35,000 to 50,000 hours of useful life, though time to complete failure may be shorter or longer.^[132] Fluorescent tubes typically are rated at about 10,000 to 25,000 hours, depending partly on the conditions of use, and incandescent light bulbs at 1,000 to 2,000 hours. Several DOE demonstrations have shown that reduced maintenance costs from this extended lifetime, rather than energy savings, is the primary factor in determining the payback period for an LED product.^[133]
• Shock resistance: LEDs, being solid-state components, are difficult to damage with external shock, unlike fluorescent and incandescent bulbs, which are fragile.^[134]
• Focus: The solid package of the LED can be designed to focus its light. Incandescent and fluorescent sources often require an external reflector to collect light and direct it in a usable manner. For larger LED packages total internal reflection (TIR) lenses are often used to the same effect. However, when large quantities of light are needed many light sources are usually deployed, which are difficult to focus or collimate towards the same target.

Disadvantages

• Temperature dependence: LED performance largely depends on the ambient temperature of the operating environment – or thermal management properties. Overdriving an LED in high ambient temperatures may result in overheating the LED package, eventually leading to device failure. An adequate heat sink is needed to maintain long life. This is especially important in automotive, medical, and military uses where devices must operate over a wide range of temperatures, which require low failure rates. Toshiba has produced LEDs with an operating temperature range of −40 to 100 °C, which suits the LEDs for both indoor and outdoor use in applications such as lamps, ceiling lighting, street lights, and floodlights.^[101]
• Voltage sensitivity: LEDs must be supplied with a voltage above their threshold voltage and a current below their rating. Current and lifetime change greatly with a small change in applied voltage. They thus require a current-regulated supply (usually just a series resistor for indicator LEDs).^[135]
• Color rendition: Most cool-white LEDs have spectra that differ significantly from a black body radiator like the sun or an incandescent light. The spike at 460 nm and dip at 500 nm can make the color of objects appear differently under cool-white LED illumination than sunlight or incandescent sources, due to metamerism,^[136] red surfaces being rendered particularly poorly by typical phosphor-based cool-white LEDs. The same is true with green surfaces.
• Area light source: Single LEDs do not approximate a point source of light giving a spherical light distribution, but rather a lambertian distribution. So, LEDs are difficult to apply to uses needing a spherical light field; however, different fields of light can be manipulated by the application of different optics or “lenses”. LEDs cannot provide divergence below a few degrees.^[137]
• Light pollution: Because white LEDs emit more short wavelength light than sources such as high-pressure sodium vapor lamps, the increased blue and green sensitivity of scotopic vision means that white LEDs used in outdoor lighting cause substantially more sky glow.^[115]
• Efficiency droop: The efficiency of LEDs decreases as the electric current increases. Heating also increases with higher currents, which compromises LED lifetime. These effects put practical limits on the current through an LED in high power applications.^[138]
• Impact on insects: LEDs are much more attractive to insects than sodium-vapor lights, so much so that there has been speculative concern about the possibility of disruption to food webs.^[139][140]
• Use in winter conditions: Since they do not give off much heat in comparison to incandescent lights, LED lights used for traffic control can have snow obscuring them, leading to accidents.^[141][142]
• Thermal runaway: Parallel strings of LEDs will not share current evenly due to the manufacturing tolerance in their forward voltage. Running two or more strings from a single current source will likely result in LED failure as the devices warm up. A circuit is required to ensure even distribution of current between parallel strands.^[143]