Lighting Research Center Lighting Research Center

Volume 3 Number 4
Copyright @1998 Rensselaer Polytechnic Institute

LEDs: From Indicators to Illumination?

What are LEDs?

Polymers Offer Glowing Prospects

Signaling the Future

Trafficking in LEDs

By Andrew Bierman

Nearly everyone is familiar with LEDs (light-emitting diodes) from their use as indicator lights and numeric displays on consumer electronic devices. The low light output of LEDs and a lack of color options have limited LED technology to these uses. Now, however, new LED materials and improved production processes have resulted in bright LEDs in colors throughout the visible spectrum with efficacies greater than incandescent lamps. These brighter, more efficacious, and colorful LEDs may move LED technology into a range of lighting applications. The energy-savings potential is enormous. By one estimate, replacing only the incandescent traffic signals in the United States alone with LED signals would save nearly 2.5 billion kilowatt hours annually (see "Trafficking in LEDs").

Exit signs change

The first LEDs bright enough to use in outdoor applications were AlGaAs (aluminum gallium arsenide). These red LEDs appeared as high-mount stop lights on automobiles and in a limited number of traffic lights. They were also used successfully in exit signs. Today the exit sign market has been almost completely transformed. A 1998 Lighting Research Center survey of exit sign sales representatives found that about 80 percent of exit signs being sold use LEDs as the primary light source. Similar transformations have occurred in roadway work zone safety lights and variable message signs when AlGaInP (aluminum gallium indium phosphide) LEDs became available. The recent advent of efficient green, blue, and white LEDs may lead to more applications.

AlGaInP and InGaN (indium gallium nitride) LEDs have succeeded AlGaAs as the brightest available LEDs.

AlGaInP LEDs range in color from red to amber and produce about 3 lumens with efficacies greater than 20 lumens per electrical watt, although green and yellow AlGaInP LEDs have much lower efficacies. Hewlett-Packard is expected to release AlGaInP LEDs with light output greater than 10 lumens per LED.

The Nichia Chemical Company in 1993 introduced InGaN LEDs with efficacies more than 100 times greater than earlier blue and green LEDs. Other companies, including Hewlett-Packard and Panasonic, offer similar InGaN LED products.

White LEDs

The previous inability of LED manufacturers to produce white LED products has been a barrier to their use in a wide range of applications.

Green InGaN LEDs have efficacies exceeding 30 lumens per watt, and blue InGaN LEDs have efficacies of 10 lumens per watt. InGaN technology also makes possible the first white LEDs.

"The goal of the semiconductor industry is to go from indicator technology to illumination technology," says Dave Evans, a Hewlett-Packard technical marketing engineer.

However, most illumination applications require white light. Manufacturers can produce white light by mixing LEDs of different colors. "Mixing discrete color LEDs produces a poor white," Evans says. He suggests that people, especially those with color-deficient vision, perceive the mixed LED white differently than they perceive a broadband white because of the narrow spectral output of the individual LEDs.

Peter Lemme, vice president of engineering at Marktech Optoelectronics, thinks that the variation of LED color properties due to manufacturing tolerances makes mixing individual LED chips impractical because special tuning would be required for each product. In addition, the various LED technologies experience different light output degradation rates that will produce color shifts over time. Moreover, he says, it would be expensive to package three chips together and mix the light well.

Alternatively, short wavelength blue LEDs can be used to activate a phosphor, which will emit a longer wavelength broadband spectrum that is perceived as white after absorbing the blue LED light. "Phosphor is absolutely better," Lemme says.

Nichia is currently the only manufacturer of white LEDs. (Hewlett-Packard and Cree Research Inc. are planning to market white LEDs soon). Since Nichia manufactures phosphors, which are used primarily in the cathode ray tube industry, they not surprisingly use a photo-active phosphor and a blue LED chip to produce white light. Nichia's white LEDs offer broadband spectrums with correlated color temperatures ranging from 4000 to 11,000 kelvin (K) and efficacies of 10 lumens per watt, says Michael Stewart, Nichia sales representative. The white LEDs are used mainly for backlighting displays and small area lighting, where they have replaced small fluorescent tubes, according to Nichia.

In Japan, effort is focused on developing efficient ultraviolet LEDs to use with phosphors to generate white light in much the same way traditional fluorescent lamps produce white light from ultraviolet.

Making the transition from small indicator lights to illumination applications requires more from the industry than white LEDs. "We learned a lesson about using AlGaAs LEDs outdoors in hostile environments," says Lemme. When first introduced, the AlGaAs LEDs in bright outdoor products such as exit signs were promoted as having 100,000+ hour lifetimes. In actual use, however, the light output of the AlGaAs LEDs decreased by about 50 percent after 15,000 to 40,000 hours.

Practically all high brightness outdoor applications now use AlGaInP technology, which is less sensitive to humidity and heat, Lemme says.

Efforts to further the use of LEDs in traffic signals focus on reducing the overall cost by reducing the number of individual LEDs per signal. Green and amber LEDs have penetrated the market much more slowly than red LEDs because of the higher cost of green LEDs and the greater number of amber LEDs needed for a signal.

Blue-green InGaN LEDs cost about $0.65 each in quantity while red AlGaInP LEDs cost only about $0.20 each. A price of $0.35 each will trigger greater use of InGaN LEDs, Evans says.

LEDs are increasingly being used as striplights for path marking and emergency wayfinding systems. Their long life and cool operation allows them to be embedded in plastic materials, which makes them perfect for these applications. For example, a flexible plastic extrusion that contains LEDs can be implanted in floors or on steps. LED modules can be fixed in roadways for lane markers that remain highly visible in rain and fog or in crosswalks that light up when a pedestrian steps onto a street.

However, the book High Brightness Light Emitting Diodes, edited by G. B. Stringfellow and M. George Craford, points out some limitations on the potential of LEDs: "The efficacies obtained are approaching their theoretical maximum values, although in the AlGaInN systems substantial further performance improvement seems plausible. There are no new materials systems evident on the horizon."

The editors speculate that at most a fivefold increase in performance is possible for AlGaInN systems, which include the InGaN blue, green, and white LEDs. That would mean that the currently available white LEDs at 10 lumens per watt could, in the future, have an efficacy of about 50 lumens per watt.

LEDs are potentially the most efficient light sources for producing small quantities of light over narrow wavelength bands. For some wavelength bands, LEDs are already the most efficient light source.

The brightest available LEDs produce about 3 lumens each. Bundling hundreds of these together to produce the light output of, say, a compact fluorescent lamp seems impractical. Even bundling a hundred or so LEDs together in a traffic-signal is expensive and introduces heating problems. High output LEDs that produce greater than 10 lumens are now being used in some traffic-signal products, significantly reducing the number of LEDs required. Current research with GaN semiconductors shows promise for larger chip sizes. There are, however, many lighting applications that require only a few lumens, or tens of lumens, for which LEDs are ideal. In the past, most of the talk about LEDs has focused on efficiency. Now that the efficiencies have exceeded other light sources, future work can focus on packaging LEDs into useful products.


What are LEDs?

LEDs are solid-state semi-conductor devices that convert electrical energy directly into light. LED "cold" generation of light leads to high efficacy because most of the energy radiates within the visible spectrum. LEDs can be extremely small and durable; they also provide longer lamp life than other sources.

Figure 1 shows a schematic of a typical T-13/4 LED. The plastic encapsulant and the lead frame occupy most of the volume. The light-generating chip is quite small (typically 0.25 millimeters square). Light is generated inside the chip, a solid crystal material, when current flows across the junctions of different materials. The composition of the materials determines the wavelength and therefore the color of light.

Figure 2, an expanded view of a blue InGaN (indium gallium nitride) semiconductor chip, shows the different layers of semiconductor material. InGaN materials produce colors ranging from blue to green.

The manufacturing of LEDs uses a process know as epitaxy in which crystalline layers of different semiconductor material are grown on top of one another. Advances in epitaxial crystal growth processes have enabled the use of LED materials for colors that previously could not be made with high enough purity and structural precision. Recent breakthroughs in the technique of chemical vapor deposition from metal organic precursors enable the cost-effective production of nitrides of the group III-metals from the periodic table including aluminum gallium indium nitrides. Highly efficient InGaN blue LEDs result from this process.

About 30 percent of the light generated inside the chip makes it way out of the brightest LEDs. Semiconductor materials have very high indices of refraction and so can trap a great deal of light when configured in a square chip. An epoxy encapsulant around the LED chip reduces the refractive index mismatch and allows more light to be emitted.

For some LEDs, the light escaping the chip (extraction efficiency) can be 4 percent or lower. Transparent substrates and thick semiconductor layers increase the extraction efficiency. Making LED chips more spherical, which is now not practical for mass production, could also significantly increase extraction efficiency.


Polymers Offer Glowing Prospects

By N. Narendran

Will light-emitting polymer (LEP) technology revolutionize the lighting and display industries? Could LEPs supplant light-emitting diodes (LEDs) in a number of seemingly secure markets and compete as an energy-saving general illumination alternative?

Some manufacturers are confident LEPs, also known as organic LEDs (OLEDs), may someday do all that and more.

Philips Electronics NV predicts that eventually "light-emitting polymers will evolve to become as flexible as fabric and thin as paper. Formed or flat, applications in the domestic, mobile, office, and public environments will place 'spread' or 'task' lighting on ceilings, walls, floors, or free hanging." Photonics Spectra (April 1997) speculates that LEPs could be used for glowing walls and flexible TV screens that roll up.

In 1990, Professor Richard Friend and Dr. Andrew Holmes at the University of Cambridge discovered that poly (p-phenylene vinylene), commonly referred to by the acronym PPV, would emit light.

In 1992, the Cavendish Laboratory at the University of Cambridge spun off Cambridge Display Technology to commercialize LEP technology. Since then, Seiko Epson, Philips Electronics, Hoechst Innovative Display Technologies Inc., UNIAX, DuPont, and Intel are some of the companies working with Cambridge Display Technology to develop commercially viable LEP products.

What are LEPs?

LEPs, which are organic semiconducting materials, and LEDs, which are inorganic semiconductors, generate light in similar ways. However, light from LEPs can be patterned like liquid crystal displays. LEPs are also thin and can be flexible.

The PPV polymer or derivatives form the active layer of most promising LEP devices. Varying the chemical composition of the PPV polymer changes its physical and electro-optical properties.

As illustrated in Figure 1, LEP displays are constructed by applying a thin PPV polymer derivative film to a glass or plastic substrate coated with a transparent indium tin oxide electrode. Some PPV polymer derivatives can be applied directly from solution, similar to manufacturing liquid crystal displays; LEDs require more sophisticated thin film deposition methods, so LEPs can be less expensive to manufacture. The polymer is sandwiched between an indium-tin oxide electrode and a metallic electrode &emdash; such as calcium, aluminum, or magnesium; it generates light when a voltage is applied.

Some LEP devices can be as bright as a cathode ray tube (around 100 candelas per square meter), with luminous efficacies between 2 to 3 lumens per watt. Researchers have been able to achieve brightness as high as 3 million candelas per square meter without heat degradation by operating LEP devices in pulsed mode, according to Cambridge Display Technology. Latest LEP device results from the company show luminous efficacies of 3 lumens per watt and 21 lumens per watt for the blue and green LEPs respectively. Cambridge Display Technology further reports that in collaboration with Seiko Epson, they have been refining the material and device design to produce devices with common architectures and emission suitable for continuous spectrum color displays.


LEP technology lends itself to the creation of ultra-thin lighting displays that will operate at a lower voltage; Philips Components estimates that LEP lighting displays will take 10 times less power than existing light emission displays. In the near term, most LEP manufacturers will produce small displays such as digital readouts on electronic devices. With further technological advances, however, the LEP industry would like to apply the technology to large area applications such as high-resolution flat panel displays.

Today, light-emitting devices with modest efficacy (3 to 4 lumens per watt) can achieve, at low voltages, brightnesses greater than those available from incandescent or fluorescent lamps. However, at these high brightness levels, the operating life is severely limited, in part due to heat generated by the high power level and the relatively low efficiency. Before lighting products will be commercially viable, therefore, efficacy must be improved to reduce heat generation when operating at high brightness. Efficacies of 25 percent or greater are theoretically achievable, and could make large-scale lighting practical.

UNIAX claims to have developed LEPs with demonstrated efficacy exceeding 3 lumens per watt and brightnesses of approximately 500 candelas per square meter at 3 volts. The company believes that achieving 25 percent efficacy would enable manufacturers to develop polymer LEP products like wallpaper lighting that would entirely change the lighting industry.

LEPs in the future

In February 1998, the first plastic television screen prototype, by Cambridge Display Technology and Seiko Epson, was unveiled; it is 50 millimeters square, black and white, and only 2 millimeters thick. The screen can be viewed from all angles and doesn't blur fast action shots, according to the manufacturers. The companies predict that they will announce full-size color TV screens soon.

Other experimental LEP prototypes that have been demonstrated are manufactured on glass substrates instead of plastic. Nick Colaneri of UNIAX said that the company and a partner hope to release 1-by-2 inch, monochrome dot-matrix displays for personal communications products as its first commercial LEP product in early 1999.

Manufacturers of LEDs and LEPs are both making significant effort towards creating continuous spectrum light or white light suitable for general illumination (see figure 2). However, Dr. S.H.A. Begemann, president of Advanced Lighting Concepts and senior vice president of Philips Lighting, says he does not foresee general lighting applications for LEPs in the coming decade.

LEP manufacturers have yet to solve the problems of their relatively low efficacy and short operating life. While progress has been made in producing white light with LEPs, problems still exist. The different emissive materials required for white-light age at different rates and create an undesirable spectral shift over time.

Many LEPs exhibit relatively high photoluminescence efficacy with minimal self-absorption. The absence of self-absorption by the polymer film is a critical advantage in creating white light. Recently, researchers have demonstrated that highly efficient white-light emission can be obtained by combining the LED and LEP technologies.

The successful development and commercialization of solid-state white light sources that are more efficacious than tungsten lamps would clearly have a major impact. These devices would have longer service life, increased heat resistance and mechanical shock resistance, higher energy efficiency, and lower costs than existing alternatives. They could have the potential to change lighting design, but whether they will become a powerful new tool remains to be seen.


Signaling the Future

By John Bullough

States and municipalities are using light-emitting diodes (LEDs) in traffic signals more frequently, citing their potential for reduced energy and maintenance costs (see "LEDs: From Indicators to Illumination?"). Pioneers include Anaheim, Denver, Philadelphia, and the state of California.

The Institute of Transportation Engineers issued an interim specification for LED traffic signals in June 1998, listing minimum performance requirements. The National Cooperative Highway Research Program is overseeing human factors research, with results expected at the end of 1999.

Questions remain about LED long-term performance and economic feasibility. Without answers, many cash-strapped agencies won't purchase LED signals. High initial costs (about $100 for a red LED signal, compared to $3 for an incandescent traffic lamp) can be difficult to justify because precise energy and maintenance savings are difficult to predict.

To find out more, transportation officials, utility representatives, manufacturers, and researchers met twice in 1998.

Roundtable Discussion

The Pacific Gas and Electric Company hosted a roundtable in San Francisco in June. Participants recommended research to confirm whether LED signals provide visibility and safety equivalent to their incandescent counterparts and dissemination of objective information about this evolving technology. In response, the Lighting Research Center (LRC) and Pacific Gas and Electric established a web site with background information, LED and signal manufacturer information, and a summary of the roundtable discussion. The address is:

With support from Hewlett-Packard, the LRC is also conducting visibility research.

Technical Workshop

The National Academy of Sciences hosted a Transportation Research Board workshop on LED technology in traffic signals in October featuring tutorials on how LEDs work, panels on the lessons learned by agencies that have tried LED signals, and presentations on visibility, economics, and the development of purchase specifications.

Nearly 100 people attended the two-day event in Irvine, California. Participants learned about funding opportunities; practical issues involved in specifying, installing, and maintaining signals; and the human factors of traffic signals. For more information, check the National Academy of Science at or the ITE website at

Trafficking in LEDs

This information is from "A Market Transformation Opportunity Assessment for LED Signals" by Margaret Suozzo of the American Council for an Energy-Efficient Economy.

Light-emitting diodes (LEDs) could save nearly 2.5 billion kilowatt hours if they replaced incandescent lamps in traffic signals. The approximately 260,000 traffic signals in the U.S. offer a large potential market for LED sources.

LEDs are essentially monochromatic and an efficient source of light for traffic signals. Incandescent lamps produce white light, and all colors except for the red, green, or yellow required must be filtered out through a lens. They also produce considerable light outside of the visible spectrum that is emitted as heat. LEDs minimize both wasted light and heat.

Switching to LED red signals alone could save nearly 1.9 billion kilowatt hours. Green LEDs could save another 350 million kilowatt hours. Retrofitting yellow incandescent lamps with LEDs can save another 65 million kilowatt hours. At the end of 1997, more than 150,000 traffic signals had been retrofitted, virtually all of them red.

Rensselaer Polytechnic Institute
LRC Intranet Web mail Lighting Research Center