Solid-State Lighting: Toward Superior Illumination

Solid-State Lighting: Toward Superior Illumination MICHAEL S. SHUR, FELLOW, IEEE, AND ARTŪRAS ŽUKAUSKAS Invited Paper Solid-state lighting technology is now emerging as a cost-competitive, energy-efficient alternative to conventional electrical lighting. We review the history of lighting, discuss the benefits and challenges of the solid-state lighting technologies, and compare two approaches for generating white light from solid-state sources based on phosphor LEDs (which could be considered as solid-state replacement of fluorescent tubes) and multichip LED lamps, which offer many advantages, such as chromaticity control, better light quality, and higher efficiency. Keywords—Color perception, injection electroluminescence, phosphor-conversion LEDs, polychromatic lamps, solid-state lighting. I. INTRODUCTION Natural light comes from the sun. The solar spectrum includes not only the visible light (with wavelengths from 0.38 to 0.78 m) but also UV light and infrared radiation. The ideal lighting source should be probably close to sunlight’s spectrum, where it overlaps with the spectrum of visual sensitivity due to retinal cone photoreceptors (see Fig. 1). Since the emergence of the artificial lighting over 500 000 years ago, and till the 20th century, it has been associated first with fire and then with hot sources of electrical light, such as an incandescent bulb. Such sources are very inefficient, Manuscript received April 5, 2005; revised May 11, 2005. The work at the Rensselaer Polytechnic Institute was supported by the National Science Foundation. The work at Vilnius University was supported by the Lithuanian State Science and Studies Foundation under COST action 529 “Efficient Lighting for the 21st Century” and the MODELITA program as well as by the SELITEC Center supported by the European Commission under Contract G5MA-CT-2002-04047). M. S. Shur is with the Department of Electrical, Computer, and Systems Engineering and Physics, Rensselaer Polytechnic Institute, Troy, NY 12180 USA (e-mail: [email protected]). A. Žukauskas is with the Institute of Materials Science and Applied Research, Vilnius University, Vilnius 10222, Lithuania (e-mail: [email protected]). Digital Object Identifier 10.1109/JPROC.2005.853537 since a high potential energy of electrons required for radiative transitions in the visible range is acquired by thermal excitation. This inevitably results in large losses of energy, i.e., such sources inherently are the sources of heat rather than of light. Also, they allow only for a limited degree of control over the radiation spectrum. Excitation of atoms and ions through acceleration of electrons also grounds the principle of operation of all discharge lamps, although some of them are deceptively considered as “cold” sources of light. For instance in one of the most efficient discharges in low-pressure mercury vapor, only 63% of energy can be converted to UV radiation. After conversion of the UV radiation to visible light in a phosphor, the overall efficiency of the resulting fluorescent lamp cannot exceed 28%. Alternative lighting devices, semiconductor LEDs, employ the effect of injection electroluminescence, which is a process of generation of light utilizing a direct conversion of electrical power to potential energy of electrons. A simplest LED is a homo- or hetero-p-n junction diode. When electric current flows through the junction, holes are injected into the depletion region and into the neutral n-type region and electrons are injected into the depletion region and into the neutral p-type region. These diffused excess electrons and holes acquire potential energy equal to the band gap energy and might recombine causing light emission. Skipping of the energy-dissipative acceleration of charge carriers in the light generation process is a revolutionary advantage of LEDs, which potentially allows attaining radiative efficiency of up to 100%. Although present commercial LEDs exhibit efficiencies less than 50% (with record efficiencies being close to 60%), their efficiencies are limited by materials quality and device-structure imperfection rather than by the basic physics. Therefore, the LED is considered as an ultimate form of lamp [1], which will gradually approach complete power-to-light conversion as technology matures. In this paper, we briefly review historical evolution of lighting technology, including key milestones and the state of the art of the development of LEDs. Then we present a 0018-9219/$20.00 © 2005 IEEE PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 1691 Fig. 1. Spectra for solar radiation at sea level when the sun is at zenith (called AM1 (Air Mass One) illumination). The power density of the AM1 radiation is 92.5 mW/cm . Thick solid line shows the relative responsivity of retinal cone photoreceptors. more detailed discussion of issues related to the quality of LED-generated white light and to the ways of attaining white light for superior-grade illumination. We conclude with the future expectations for the development of solid-state lighting. Additional information about the solid-state lighting in general and about LEDs in particular can be found in recent books [2]–[5], review papers [6], [46], and on the Web [7]–[9]. II. BRIEF HISTORY OF LIGHTING1 Up to the beginning of the 19th century, artificial light was produced by pyroluminescence, which is due to radiative transitions in excited atoms and ions, recombination of ions to form molecules, and incandescence of solid particles in the flame. This pyroluminescence-based lighting dates as far as 500 000 years ago, and we can probably argue that the use of artificial lighting was a prerequisite for human civilization. For thousands of years, braziers (dishes where fire occurs), cressets (fire baskets on poles), and torches (made of vegetables treated with flammable substances—pitch, wax, resin, tallow, oil, etc.) were the only artificial sources of light. Stone lamps with a wick—a capillary cord that draws fuel up to a flame appeared 30 000 to 70 000 years ago, according to archeological findings and cave paintings. Later, lamps made of shell, pottery, and metal appeared, but without substantial improvement in the performance of the emission process. Candles, which appeared in Roman times, also contain a wick but with the fuel (such as beeswax, tallow, and later, paraffin wax) melting in the heat of the flame. Modern lighting-device engineering dates back to the end of the eighteenth century, when A. L. Lavoisier discovered that combustion is due to oxygen in the air. Using this knowledge, Ami Agrand of Geneva invented an oil lamp with a tubular wick placed within two concentric tubes and a glass chimney around the burner. He achieved a tenfold gain in 1This 1692 section is based in large part on an historical overview given in [3]. light due to a better supply of air to the flame and the resulting higher burning efficiency. The lamp was demonstrated to King George III, and Agrand was granted an English patent (no. 1425 of 1784). Scottish inventor William Murdoch introduced gas lighting in 1772. This kind of lighting rapidly became the mainstream technology of the civilized world only to be replaced by electric lighting. In 1826, Thomas Drummond invented the first crystalline lighting device (called the limelight) based on candoluminescence, emission in excess of blackbody incandescence due to thermal excitation of ions (discovered in 1820 by Goldsworthy Gurney). The device consisted of a cylinder of lime (calcium oxide), which was brought to a state of dazzling brilliancy by the flame of an oxyhydrogen blowpipe. Limelight was used in theaters in the 1860s and 1870s until superseded by the electric arc. The history of electric lighting goes back all the way to the 17th century, when the effect of the luminous discharge of static electricity in mercury vapor was discovered. But only in the beginning of the 19th century, Sir Humphrey Davy demonstrated discharge between two rods of carbon (an arc) and the glowing of a piece of wire heated by electric current (incandescence). A copper–zinc battery invented by Alessandro Volta in 1800 was used as a power source. However, the true change from flame to electric in lighting happened only in the 1870s, when Z. T. Gramme introduced an efficient continuous-current generator (dynamo machine) and Paul Jablochkoff (Pavel Yablochkov) demonstrated the first practical electric lighting device in 1876. Despite a small lifetime (just a few hours), Jablochkoff’s candles were immediately adapted for street illumination. In a few years, they were replaced by higher performance, longer lifetime (up to 1000 hours) carbon-arc devices. Such discharge lamps were used widely for street lighting up to the second decade of the 20th century and for aircraft floodlighting during both world wars. Thomas Alva Edison and Joseph Wilson Swan were recognized as the inventors of the incandescent filament lamp PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 Fig. 2. Replica of Edison bulb. through famous patent trials. Swan demonstrated an incandescent lamp earlier in 1879; however, the first topics of his patients were the methods of evacuation and prevention of fracture of the glass at seals (British patents nos. 18 and 250 of 1880). Edison demonstrated his device a couple of months later, but he obtained a patent for a filament lamp (U.S. patent no. 223 898 of 1879; see Fig. 2). The companies that promoted the incandescent lamp now lead the electric and lighting industry: General Electric Company (GE, descendant of the Edison Electric Light Company), (British) General Electric Company (GEC), AEG, Siemens, Osram, and Philips.). In 1900, Peter Cooper Hewitt patented the mercury vapor lamp, and in 1938, GE and Westinghouse Electric Corporation introduced new colored and white low-pressure mercury discharge lamps with the inside of the tube coated with a phosphor powder. The fluorescent lamp employs photoluminescence excited by UV emission of mercury and is much more efficient than the incandescent lamp. At present, tungsten incandescent lamps provide most of residential lighting. Lighting in office, industrial, and commercial establishments rely on fluorescent lamps and metal–halide lamps, which utilize discharge in high-pressure mercury vapor with supplementary metals introduced using halides. Cheap light from high-pressure sodium lamps is used for street illumination. At the time when crystal-based limelight was almost forgotten and the lighting community was excited by the invention of the tungsten filament by A. Just and F. Hanaman (1903), H. J. Round discovered electroluminescence in a silicon carbide semiconductor. His short communication dated 1907 (see Fig. 3) describes an observation of yellow and even blue light when a bias was applied to silicon carbide crystallite through a metal needle. Unfortunately, his explanation of this phenomenon involving thermoelectric effects was completely wrong. It took about 60 years before N. Holonyak, Jr. and S. F. Bevacqua developed a p-n junction emitter of visible (red) light [10]. LEDs with the efficiency exceeding that of filtered incandescent lamp (high-brightness LEDs) were developed based on double heterostructures proposed Fig. 3. The first paper on injection electroluminescence. Fig. 4. Pioneers of electrical lighting. by Kroemer [11]. Interestingly enough, the development of both visible emitter and double heterostructure was primarily targeted at improving of semiconductor lasers. Current highbrightness LEDs rely on three III-V heterostructure materials systems, which are AlGaAs, AlGaInP, and AlInGaN. The most recent development of the nitride materials system yielded blue and near-UV LEDs, which made possible generation of white light. This development is based on the pioneering work of J. I. Pankove and Maruska in the 1970s and a breakthrough in developing of candela-class blue LEDs by S. Nakamura and co-workers in 1994 [2]. Fig. 4 shows pioneers of electrical lighting. Fig. 5(a) and (b) shows the current status for luminous efficiency and external quantum efficiency for two main families of high-brightness LEDs—AlInGaN and AlGaInP SHUR AND ŽUKAUSKAS: SOLID-STATE LIGHTING: TOWARD SUPERIOR ILLUMINATION 1693 Fig. 5. (a) Luminous efficiency and (b) external quantum efficiency for two families of LEDs—AlInGaN and AlInGaP based. Arrows show luminous efficiencies for incandescent bulbs (about 14 lm/W) and fluorescent tubes (78 lm/W). Horizontal bars represent the present and expected performance of white LEDs. Solid lines show the human eye photopic sensitivity. Most of the data are from M. G. Craford, presented at the 2004 International Technology Conference, Feb. 2004, Hong Kong, and at the Nanoscience and Solid State Lighting Department of Energy Nanosummit, June 2004, Washington, DC. Fig. 6. (a) Schematic structure of a dichromatic phosphor-conversion white LED. (b) CIE 1931 color mixing diagram with the chromaticity coordinates of the emission from the blue InGaN LED and YAG : Ce phosphor (filled dots) and the resulting white light (star). based. (Red AlGaAs LEDs are gradually superseded by the AlGaInP LEDs, which have a higher efficacy of radiation due to shorter “red” wavelengths available.) Also shown are luminous efficiencies for incandescent bulbs and fluorescent tubes. As seen, the performance of yellow LEDs falls short, which is significant for the performance of multichip white LEDs discussed below. Further progress is expected to come due to the design improvements leading to a higher quantum efficiency and better light extraction using, for example, photonic crystals. However, the most promising expectations for solid-state lighting technology rely on developing novel sources of white light. Two different approaches considered below compete for applications in white LEDs—phosphor-conversion LED lamps and multichip polychromatic LED lamps. III. PHOSPHOR-CONVERSION WHITE LEDS Due to peculiarities of radiative recombination in semiconductors and because of low temperature of electrons and holes in the active layers of semiconductor structures, 1694 the LED light is emitted within relatively narrow-band spectra. Therefore, LEDs are inherently colored sources of light, whereas the most promising applications of solid-state lighting technology require white light with a broad spectrum (see Fig. 1). Two basic approaches for producing white light rely on either partial or complete conversion of short-wave radiation from semiconductor chips in phosphors and on using of a variety of independently controlled primary colored LEDs, respectively. The first approach leads to a solid-state replacement of the fluorescent lamp and offers relatively cheap single-chip devices, which already found numerous applications in low-power, high-reliability, and compact lighting fixtures. The second approach is a revolutionary technology with completely new possibilities in lighting, such as dynamic control of color, a tradeoff between the efficiency and quality of illumination, feedback-stabilized parameters of light, and superior visual performance. Schematic structure of a conventional yellow–blue phosphor-conversion LED is shown in Fig. 6(a) [2], [12]. A part of the blue radiation emitted by the InGaN chip is PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 absorbed in the phosphor particles dispersed in a transparent resin and is converted to yellow light. The remaining part of the blue radiation escapes to the ambience. The mixture of these two components is perceived as white light. Fig. 6(b) shows the Commision Internationale de l’Éclairage (CIE) 1931 chromaticity diagram with the locus of chromaticities of the blackbody radiator (Planckian locus). Filled dots mark the chromaticity coordinates of the emission from the InGaN chip and cerium-doped yttrium–aluminum garnet YAG Ce . The dashed line comprises the chromaticity coordinates available by mixing of the emission from these two primary sources. For an appropriate proportion of the primary fluxes, white light with the chromaticity at the intersection of the dashed line with the Planckian locus (marked by a star) can be obtained. The emission and the excitation spectra of the garnet can be precisely tailored by substitution for Y and Al atoms by Gd and Ga atoms, respectively. Therefore, an optimized Y Gd Al Ga O Ce phosphor system is typically used. Commercial white phosphor-conversion LEDs are already at least twice as efficient as incandescent bulbs. Also, the price of LED light is rapidly approaching that of conventional lamps (see Section V). A constant increase in efficiency and decrease in price [13] makes solid-state lighting technology attractive for mass applications, such as general lighting. One of the remaining hurdles for further penetration of solid-state lighting technology is the quality of light delivered by white LEDs. The quality of white light utilized for illumination of various objects implies the ability to properly render the color of the objects. The present industry standard for color-rendering properties of light relies on color rendering indexes (CRIs), which rate the color difference of particular test samples under the source under test compared to a reference source (usually, a blackbody radiator or a specified “daylight” illuminant). The rating algorithm (test-color method) introduced more than 40 years ago [14] experienced only minor modifications since that time [15]. A special CRI for an th test sample is specified through the color shift in the CIE 1964 uniform color space (1) where , , and are the differences in chromatand , and lightness index , respecicness indexes tively, with chromatic adaptation of the visual system taken into account. The lightness index measures the ability of the sample to reflect luminous flux, while the chromaticness indexes are derived from the chromaticity coordinates and the lighting index. Fourteen test samples are specified for the assessment of the color rendering properties of light sources, and eight of those are averaged for determining an integral . Although the algorithm figure of merit, the general CRI fuses the separate differences in hue, saturation, and lightness of the samples and employs only a small fraction of possible samples, it is widely accepted for rating of sources of light. Fig. 7. Spectral power distributions for: (a) the standard phosphor-conversion LED and (b) a complimentary lamp composed of the white, red, and cyan LEDs. Dashed line shows the spectral power distribution of the CIE illuminant C (“overcast daylight”) (after [17]). To estimate the special and the general CRIs, the spectral power distribution of light emitted by a source under consideration is required. The solid line in Fig. 7(a) shows the emission spectrum of the conventional InGaN/YAG Ce phosphor-conversion LED. The spectrum contains a blue component due to electron-hole recombination in the semiconductor structure and a broad yellow band due to radiative transitions in Ce ions. The spectrum corresponds to the correlated color temperature of 6725 K, which is close to that of the CIE standard illuminant C “overcast sunlight” (6770 K correlated color temperature). However, as seen, the LED spectrum is considerably different from that of the “overcast sunlight” illuminant shown by dashed line in Fig. 7(a). Neglecting the eye-insensitive edges of the spectrum, the main difference is in the lack of power in the blue–cyan (around 470 nm) and the red (below 600 nm) regions. As a result, the general CRI amounts only to 73 points, with poor rendering of many colors. Gray bars in Fig. 8 show CRIs for the conventional white LED. The minimal value of the special indexes is for the , while the “strong yellow” “strong red” sample and “strong blue” samples also suffer from poor color rendering. Although conventional blue–yellow LEDs have already caught up with fluorescent lamps in terms of efficiency (78 lm/W [16]), inherently poor color rendering properties of this solid-state system limits their use to niche applications, such as signal lights, flash lamps, and LCD backlighting. Similar to its gas-discharge counterpart, which is the fluorescent lamp, the InGaN/YAG Ce phosphor-conversion LED cannot compete for applications requiring high quality of light, such as residential and medical lighting. The dichromatic white LED with partial conversion can be improved by supplementing it with red–orange and cyancolored LEDs [17]. Fig. 7(b) shows the emission spectra of such a lamp using supplementary LEDs. The lamp features a reduced color temperature. However, the spectral dip at about 480 nm still persists because of the color-mixing constraints SHUR AND ŽUKAUSKAS: SOLID-STATE LIGHTING: TOWARD SUPERIOR ILLUMINATION 1695 Fig. 8. CRIs for the standard phosphor-conversion LED (gray bars) and the complementary lamp (white bars) (after [17]). for the selection of the supplementary LEDs. Nevertheless, the general CRI increased up to 90 points, and the minimal value of the special CRIs amounted to 77 points (white bars in Fig. 8). Further evolution of single-chip phosphor-conversion LEDs targets several issues of light quality. One of those is a lower color temperature needed for designing solid-state lamps similar to widely used incandescent bulbs. Low color temperatures (around 3000 K) cannot be attained using only YAG Ce -based phosphors, since the emission spectrum due to cerium ions feature too small values of the chromaticity coordinate; see Fig. 6(b). Another direction is to improve the color rendering properties. This can be achieved by generating a smoother spectrum by using of multiphosphor blends and near-UV or even deep UV LEDs. In such white LEDs, the initial near-UV radiation from the semiconductor structure is completely converted in phosphors, and the resulting spectrum does not contain narrow-band lines. New phosphor systems for white LEDs are being developed to meet the above requirements for light quality. One of the first alternative phosphor families proposed is chalcogenides ZnS : Ag (blue), ZnS : Cu, Al (green), and ZnCdS : Ag (red), which can be excited by near-UV LEDs [18]. More sophisticated rear-earth (Eu) doped chalcogenide phosphors were considered for white LEDs with a special purpose to enhance the red emission. In particular, a white LED with partial conversion of blue light in SrGa S Eu (green) and SrS Eu (red) was proposed with the general CRI values exceeding 90 points [19]. Another option is to Eu for green but ZnCdS : Ag,Cl for red use SrGa S emission [20]. A bicolor (green–red) CaGa S CaS Eu phosphor for excitation by a blue LED was developed and exhibited both higher efficiency and better color rendering than the YAG Ce counterpart [21]. A recent 1696 addition to chalcogenide phosphors for white LEDs is the Ca Sr Se Eu phosphor [22], which emits in the orange to yellow–green region depending on the Sr molar fraction. Silicates are the extension of the oxide phosphor family already represented by yttrium–aluminum garnets. Silicates are more stable against UV radiation and less toxic than chalcogenides. Yellow Sr SiO Eu and Sr SiO Eu silicate phosphors were developed to substitute for YAG Ce [23], [24]. These phosphors exhibited higher efficiency than the Ce-doped garnet. However, the resulting white LEDs had no improvement in the general CRI. An innovative approach is to use Ba MgSi O Eu , Mn [25] and Sr MgSi O Eu , Mn [26] tricolor phosphors (blue–cyan–red and blue–yellow–red, respectively), which can be excited by near-UV LEDs. The phosphors emit two short-wavelength bands due to the radiative transitions in ions and the red band originating from the Mn Eu ions. The resulting white LEDs exhibited the general CRI of 85 and 92 points, respectively, which is considerably Ce LEDs. higher than for commercial InGaN/YAG Other promising oxide phosphors for near-UV excitation are EuAlO Eu (red), EuAlO Eu (green), Y SiO Ce (blue), and Y SiO Ce , Tb (multicolor white) [27]. Ru-doped BaMgAl O (blue) and Y O (red) combined with SrGa S Ru (green) were also employed in white LEDs using near-UV excitation [28]. Recently, highly stable red phosphors based on nitride ceramics have been developed. A europium activated strontium–calcium–silicon–nitride (SCESN) phosphor emitting at about 660 nm was used to enhance the long-wavelength emission in the conventional InGaN/YAG Ce system [29]. The color temperature dropped to 4670 K, and the general CRI increased to 88 points. The major improvement PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 in color rendering was achieved for the “strong red” special that attained values in excess of 60 points. Another CRI example is an oxynitride phosphor Ca SiAlON Eu [30]. This phosphor emits at about 600 nm, which allows for the development of white LED with low color temperatures using a partial conversion of the blue light. Nevertheless, phosphor-conversion white LEDs suffer from the same drawbacks as their ancestors, fluorescent lamps. First, a portion of energy is inherently wasted in the downconversion process. Second, phosphors offer insufficient versatility in spectral composition of light for attaining the best illumination quality. Finally, the nonuniform temperature and aging drifts in efficiency of the semiconductor chip and phosphors result in uncontrollable chromaticity shifts. This effect can reduce the “color lifetime” of the LEDs to well below the device lifetime. IV. MULTICHIP POLYCHROMATIC WHITE LAMPS Solid-state white lamps composed of independently controlled primary sources offer new opportunities for lighting technology. Since such lamps do not utilize the downconversion process, they can be superior in efficiency compared to their phosphor-conversion counterparts. However, the main advantage of multichip lamps is the possibility of controlling colors and stabilizing the chromaticity and the output flux against the nonuniform thermal and aging drifts of the primary sources. Semiconductor technology is much more versatile than phosphors technology in terms of tailoring the wavelengths of the primary emitters. As seen from Fig. 5, the peak LED wavelength can be varied within a wide range by using different semiconductor alloy compositions for the active region of the light-emitting structure. This allows for the fabrication of the primary sources depending on the required properties of a multichip lamp and trading off and optimizing efficiency and color rendering. The solution of the optimization problem yields the number and the peak wavelengths of the primary LEDs as well as the relative fluxes for each group of the LEDs [3]. A general approach for optimization of multichip solid-state lamps composed of an arbitrary number of primary LEDs has been developed based on a stochastic method [31]–[33]. The optimization routine is targeted to maximize an appropriate objective function. For a tradeoff (which is the luminous between the radiation efficacy flux per unit power radiated for a particular spectrum) and the general CRI , the objective function is given by (2) where is the weight that controls the tradeoff between the , and and are efficacy and the general CRI the peak wavelengths and the relative fluxes of the primary sources, respectively. The radiation efficacy and the general Fig. 9. (a) Peak wavelength selection chart for white lamps composed of two, three, four, and five primary LEDs. The lamps are optimized for the highest luminous efficacy at a color temperature of 4870 K (after [31].) (b)–(e) Examples of spectral power distributions of the white lamps for various numbers of the primary LEDs (after [32]). CRI are not the only possible figures of merit. Other quantities like the device efficiency or cost of light might be also involved in the optimization procedure. , the solutions can For two primary sources be obtained by simply running over one of the wavelengths and finding the complementary wavelength from the color-mixing equations. The optimal solutions for the 30-nm-wide emission lines of the primary LEDs and the color temperature of 4870 K of the resulting white light are shown by solid triangles in Fig. 9(a). Lower values of the general CRI correspond to higher efficacies. The highest efficacy is obtained for the 452/571-nm system at [the corresponding spectrum is presented in Fig. 9(b)]. The general CRI can be increased up to 20 points at an expense of efficacy by shifting the primary peaks toward longer wavelengths. , a trichromatic system is required For [open circles in Fig. 9(a)]. At a reasonable value of [Fig. 9(c)], the wavelength triad is 459/537/604 nm, which is close to the optimal solution for trichromatic phosphor lamp (450/540/610 nm) [34]. For the general CRI values in or quintichroexcess of 85 points, quadrichromatic solutions emerge. The optimal quadrichromatic matic system with requires the peak wavelengths at 454/509/561/619 nm, while the general CRI of 99 points can be achieved in a quintichromatic system (448/493/531/ 572/623 nm); see Fig. 9(d) and (e). A detailed analysis of the special CRIs (Fig. 10) shows that the dichromatic system is unsuitable for illumination when discrimination of colors is important (most of the SHUR AND ŽUKAUSKAS: SOLID-STATE LIGHTING: TOWARD SUPERIOR ILLUMINATION 1697 Fig. 10. CRIs of the white lamps with the spectral power distributions shown in Fig. 9(b)–(e). special CRIs have negative values). The LED-based trichromatic system shows a very poor rendering for the “strong and only moderate values for some other red” sample samples . This makes solid-state RGB lamps uncompetitive in respect to the incandescent lamp. In contrast, the quadrichromatic and quintichromatic systems feature even distributions of the special CRIs, with all the values being in excess of 83 and 94 points, respectively. Based on this analysis, one can conclude that quadrichromatic and quintichromatic lamps can meet all the needs in high-quality lighting. However, to promote these sources for such applications as residential or medical lighting, the generated light might require quality that is beyond that rated by the standard test-color method. Below, an alternative rating procedure based on subjective psychophysical evaluation will be demonstrated. At present, polychromatic solid-state lamps can rely only on a limited set of peak LED wavelengths that are manufactured mostly for full-color video displays and traffic/automotive signage. A quadrichromatic solid-state lamp based on commercially available high power LEDs was developed [35], [36]. A set of the primary LEDs with the peak wavelengths of 441 nm (blue), 523 nm (green), 594 nm (amber), and 638 nm (red) was selected by searching in the efficiencygeneral CRI space and the relative fluxes from each group of . LEDs were optimized to achieve the highest values of At a color temperature of 2600 K, the red–amber–green–blue (RAGB) lamp featured the general CRI of 86 points, which is considerably lower than the value for the optimal set of primary wavelengths (about 98 points). A standard CRI analand ysis revealed a deficiency in “strong red” a moderate rendering of “light reddish purple” with the rest special CRI all being in excess of 80 points. The stability analysis of the RAGB lamp [37] showed an intolerable variation of the color temperature and general CRI due to dependence of the fluxes delivered by red, amber, and 1698 Fig. 11. Block diagram of a stabilized quadrichromatic solid-state lamp (after [35]). green LEDs on their junction temperature. Meanwhile, technological deviations of the peak wavelength due to a spread in semiconductor alloy composition were found to be of marginal importance. To stabilize the lamp against the temperature and aging drifts, the driving circuit was equipped by a digital feedback system [38]. Fig. 11 shows the block diagram of the stabilized lamp. The primary LEDs contained in the fixture are driven using the pulse-width modulation technique to prevent chromaticity drift due to driving current variation. The pulses are formed by current regulators, which are digitally controlled by a microcontroller. The microcontroller is preset by a remote computer for a particular combination of the primary fluxes required for the selected color temperature and overall flux. The lighting fixture is equipped by a photodiode sensor. The signal provided by the sensor is digitized in the microcontroller unit and employed to adjust the duration of the current pulses using initial calibration data. The PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 Fig. 12. Segment of the CIE 1976 (u ; v ) color plane with the chromaticity coordinates of 40 Munsell samples. Open circles and triangles are physical chromaticities under tungsten and RAGB lamps, respectively; filled circles and triangles are corresponding subjective chromaticities. Crosses show the calculated chromaticities under standard illuminant C “overcast daylight.” The angular position of two dashed vectors define the subjective hues of a Munsell sample with the dominant wavelength of 485 nm (encircled points) obtained under two sources of light, tungsten and solid-state RAGB lamps;  is the hue difference (after [36]). Fig. 13. (a) Calculated and (b) subjective relative differences in hue for the Munsell samples under RAGB versus tungsten lamps. (c) Spectrum of the RAGB lamp at 2600 K color temperature. (d) Normalized spectral sensitivity of tree types of retinal cones (after [36]). measurement and adjustment procedure is being performed periodically without interruption of the driving process. The above-described stabilized multichip RAGB lamp was used in a psychophysical experiment on subjective perception of colors [36]. Hue distortions, which are known to be the most important for subjective rating of perceived colors [39], were investigated using 40 samples from the Munsell palette (value 6, chroma/6, and hue incremented by 2.5). Subjects with normal trichromatic color vision were asked to match the color of the stimuli displayed on a calibrated monitor with that of the sample illuminated by either a tungsten lamp or the RAGB lamp with the same color temperature (2600 K). The acquired settings of the monitor were used to extract the perceived chromaticities of the samples. Fig. 12 shows the results of the psychophysical investigation for one of the subjects. The data are plotted color plane, where the distances over the CIE1976 between chromaticity points are well-matched to the subjective discrimination of colors. The open circles and triangles show the calculated (physical) chromaticities of the samples under the tungsten and RAGB lamps, respectively. The filled circles and triangles show the subjectively perceived chromaticities under the same sources. The loci of the perceived chromaticities are seen to be shifted quite significantly with respect to the corresponding physical ones toward the locus of chromaticities calculated for standard illuminant C (“overcast daylight”; crosses in Fig. 12). This effect of color constancy, which is a property of human vision to recognize colors under different sources of light as being the same, varies for different subjects and is not accounted in the standard test-color method. Nevertheless, each sample illuminated by a particular source can be characterized by a perceived hue, which is the angular position of a vector connecting the chromaticity point of the source and the point of the perceived chromaticity (dashed lines in Fig. 12). A tested source of light (the RAGB lamp) can be rated with respect to the reference source (such as the tungsten lamp) , as shown in Fig. 12. by the difference in hue Fig. 13(a) and (b) shows the physical and subjective (averaged over seven subjects) hue distortions that occur under illumination by the stabilized RAGB lamp as compared to the tungsten lamp. The hue differences for each sample are normalized to the hue discrimination angle, which is the hue difference between the neighboring samples. Subjective hue distortions [Fig. 13(b)] are somewhat larger than the physical ones [Fig. 13(a)] due to the above-mentioned effect of the color constancy and due to experimental uncertainties. Actually, a relative difference in hue that exceeds unity means that the subject perceives the wrong hue. However, the striking result of this study is that significant hue distortions are concentrated in the cyan (around 490 nm) and yellow (around 580 nm) regions that are considered as properly rendered in the standard CRI analysis. These critical regions correspond to a deficiency of power or a strong variation of the power in the RAGB lamp spectrum [Fig. 13(c)]. On the other hand, they exactly match the overlap regions of the sensitivity functions of different retinal cone photoreceptors [Fig. 13(d)], where color discrimination is high due to large differences in receptor excitements [40]. The critical cyan region falls in between the slopes of the short-wavelength (S-type) and middle-wavelength (M-type) cones, while the critical yellow region is on the slope of the M-type cone sensitivity function overlapped with the long-wavelength (L-type) photoreceptor function. For practical applications, the yellow region is of major importance for subjective rating of lighting in residential and other social environments. The results of the above-described tentative psychophysical experiments imply that rating of light sources based on the standard test-color method can be improved and that it is possible to promote the quality of polychromatic sources of SHUR AND ŽUKAUSKAS: SOLID-STATE LIGHTING: TOWARD SUPERIOR ILLUMINATION 1699 Fig. 14. Luminous efficiency (lm/W) for different lighting sources. The insert shows the projected trend for white LEDs with CRI of at least 80. Fig. 15. Projected cost of light. Arrows show the present cost of ownership for incandescent and fluorescent lighting. Data from [6]. . white light beyond the well-established limit of In particular, the primary color selection is to be extended for proper perception of hues in the cyan and yellow critical regions of dominant wavelengths. To this end, multichip polychromatic solid-state lighting technology, which offers unsurpassable versatility in selection of the primary wavelengths, should avoid compulsive drawbacks of fluorescent lamps and phosphor-conversion LEDs. Solid-state lighting technology is capable of delivering efficient white sources that can be comparable in quality of light with energy-wasting Planckian radiators (incandescent lamps) and natural daylight. For instance, quintichromatic solid-state lamps, which contain primary sources in both the conventional RGB regions and critical cyan and yellow regions, might be an ultimate solution. Such lamps can be optimized not only for high values of the general and special CRIs but also for small distortions of hues. Current mathematical methods and computer capabilities allow one to perform such an optimization for hundreds of color samples. However, practical implementation of the optimal quintichromatic and higher order lamps is still an issue for several reasons. First, there is no industry standards for LED wavelengths that meet the needs of the polychromatic sources. The marketed tricolor RGB modules are optimized for a wide color gamut required by the TV standards rather than for generation of illumination-grade white light. Second, LEDs 1700 with high radiant efficiency are available only at the edges of the visual spectrum (see Fig. 5). Such are InGaN LEDs in the blue region [16], [41] and AlGaInP LEDs in the red to orange region [42]. The efficiency of nitride LEDs rapidly decreases while moving to longer (green) wavelengths because of the materials quality issues of the InGaN alloy with a higher In molar fraction. At the same time, the efficiency of AlGaInP LEDs drops while moving to shorter wavelengths because of the alloy band structure approaches the direct-toindirect band crossover. As a result, there are no efficient inorganic-semiconductor emitters for the yellow–green region (560–580 nm), where the sensitivity of the human vision is highest. This shortcoming can be resolved only by using InGaN-based colored LEDs with a complete phosphor conversion [43] or by introduction of reliable organic LEDs [44]. V. PROJECTIONS FOR THE FUTURE Although improvement of the quality of light from LEDs is an important issue, most experts link the future penetration of solid-state lighting technology with such quantitative figures of merit as LED efficiency and cost. In Fig. 14, typical luminous efficiencies of conventional light sources are collated with the present efficiency of red LEDs and present and projected efficiencies of white LEDs. The main difference between the conventional sources and LEDs is in that PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 specific needs—from general to medical and agricultural lighting, lighting for space flights, lighting for elderly, lighting for people with special color needs, lighting for animals, lighting for museums and for illuminating art objects, and lights for theatrical productions [45]. Low-voltage driving, fast switching, and compatibility with networked computer controls enable intelligent lighting systems with software-controlled stability, operating function, adaptation, and energy saving. Such systems are expected to emerge and become disruptive and revolutionary technology in the near future. REFERENCES Fig. 16. Expected penetration of LED lighting and related cost savings. Data from R. Haitz, F. Kish, J. Tsao, and J. Nelson, “Innovation in semiconductor illumination: Opportunities for national impact,” 2000. the former have almost reached their physical limit of efficiency, while the latter have not (see projections in the insert in Fig. 14). For some colors, LEDs are already the most efficient artificial sources of light. Based on the physical principles that are already known, the luminous efficiency of white LEDs, both phosphor-conversion and multichip polychromatic, can approach 200 lm/W by the year 2020 (see the insert in Fig. 14). This corresponds to approximately 60% to 70% radiant efficiency, which can be achieved by further reduction of nonradiative recombination by improving the materials quality and designing novel structures with enhanced light-extraction efficiency. The cost of light offered by solid-state general lighting technology is now approaching the cost of light generated by the most expensive conventional technology based on the incandescent bulbs. However, solid-state general lighting is expected to become price competitive with such sources of cheap light as fluorescent lamps quite soon (see Fig. 15). Although these cost predictions primarily relate to the phosphor-conversion LEDs, a further decrease in price of the colored LEDs and the development of specialized control electronics will make multichip polychromatic sources of white light cost efficient as well. Emerging solid-state lighting technology promises tremendous cost savings even at a relatively modest penetration of LEDs into the lighting market, as shown in Fig. 16. However, solid-state lighting is capable of capturing even a larger segment of the general lighting market than shown in Fig. 16, provided that a price breakthrough will take place in the near future. LEDs have changed the concept of lighting not only in an expectation of the ultimate efficiency but also in tremendous opportunities for versatile and “smart” lighting applications. LED-based lighting technology, solid-state lighting [3], allows for unprecedented versatility in control over the radiation spectrum, which can be tailored for [1] N. Holonyak Jr., “Is the light emitting diode (LED) an ultimate lamp?,” Amer. J. Phys., vol. 68, pp. 864–866, Sep. 2000. [2] S. Nakamura and G. Fasol, The Blue Laser Diode: GaN Based Light Emitters and Lasers. Berlin, Germany: Springer-Verlag, 1997. [3] A. Žukauskas, M. S. Shur, and R. Gaska, Introduction to SolidState Lighting. New York: Wiley, 2002. [4] E. F. Schubert, Light-Emitting Diodes. Cambridge, U.K.: Cambridge Univ. Press, 2003. [5] UV Solid-State Light Emitters and Detectors ser. Proc. NATO ARW, Series II M. S. Shur and A. Žukauskas Eds. Dordrecht, The Netherlands: Kluwer, 2004, vol. 144. [6] J. Tsao, “Solid-state lighting: lamps, chips and materials for tomorrow,” IEEE Circuits Devices Mag., vol. 20, pp. 28–37, May/ Jun. 2004. [7] A. Zukausaks, M. Shur, and R. Gaska, “Solid state lighting,” [Online]. Available: http://nina.ecse.rpi.edu/shur/ Solid%20State%20LightingTutorial_F2001.pdf [8] “Solid state lighting: Overview,” [Online]. Available: http:// lighting.sandia.gov/Xlightingoverview.htm [9] Light emitting diodes.org home page. [Online]. Available: http:// www.ecse.rpi.edu/~schubert/Light-Emitting-Diodes-dot-org/ [10] N. Holonyak Jr. and S. F. Bevacqua, “Coherent (visible) light P )junctions,” Appl. Phys. Lett., vol. 1, emission from Ga(As pp. 82–83, Dec. 1962. [11] H. Kroemer, “A proposed class of heterojunction injection lasers,” Proc. IEEE, vol. 51, pp. 1782–1783, Dec. 1963. [12] P. Schlotter, R. Schmidt, and J. Schneider, “Luminescence conversion of blue light emitting diodes,” Appl. Phys. A, vol. 64, pp. 417–418, Apr. 1997. [13] D. A. Steigerwald, J. C. Bhat, D. Collins, R. M. Fletcher, M. O. Holcomb, M. J. Ludowise, P. S. Martin, and S. L. Rudaz, “Illumination with solid state lighting technology,” IEEE J. Sel. Topics Quantum Electron., vol. 8, no. 2, pp. 310–320, Mar./Apr. 2002. [14] D. Nickerson and C. W. Jerome, “Color rendering of light sources: CIE method of specification and its application,” Illum. Eng., vol. 60, pp. 262–271, Apr. 1965. [15] Bureau Central de la CIE, “Method of measuring and specifying color rendering properties of light sources,” Paris, France, Pub. 13.3, 1995. [16] J. Edmond, A. Abare, M. Bergman, J. Bharathan, K. L. Bunker, D. Emerson, K. Haberern, J. Ibbetson, M. Leung, P. Russel, and D. Slater, “High efficiency GaN- based LEDs and lasers on SiC,” J. Cryst. Growth, vol. 272, pp. 242–250, Dec. 2004. [17] P. Vitta, A. Zukauskas, R. Gaska, and M. S. Shur, “White complementary solid-state lamp,” Leukos, vol. 1, pp. 59–66, Jul. 2004. [18] Y. Sato, N. Takahashi, and S. Sato, “Full-color fluorescent display devices using a near-UV light-emitting diode,” Jpn. J. Appl. Phys., vol. 35, pt. 2, pp. L838–L839, Jul. 1996. [19] R. Mueller-Mach and G. O. Mueller, “White light emitting diodes for illumination,” Proc. SPIE, vol. 3938, pp. 30–41, Apr. 2000. [20] Y. D. Huh, J. H. Shim, Y. Kim, and Y. R. Do, “Optical properties of three-band white light emitting diodes,” J. Electrochem. Soc., vol. 150, pp. H57–H60, Feb. 2003. [21] J. Zhang, M. Takahashi, Y. Tokuda, and T. Yoko, “Preparation of Eu-doped CaGa S CaS composite bicolor phosphor for white light emitting diode,” J. Ceram. Soc. Jpn., vol. 112, pp. 511–513, Sep. 2004. SHUR AND ŽUKAUSKAS: SOLID-STATE LIGHTING: TOWARD SUPERIOR ILLUMINATION 0 1701 [22] X. Zhang, L. Liang, J. Zhang, and Q. Su, “Luminescence propSr )Se : Eu phosphors for white LEDs applicaerties of (Ca tion,” Mater. Lett., vol. 59, pp. 749–753, Mar. 2005. [23] J. K. Park, M. A. Lim, C. H. Kim, H. D. Park, J. T. Park, and S. Y. Choi, “White light-emitting diodes of GaN-based Sr SiO : Eu and the luminescent properties,” Appl. Phys. Lett., vol. 82, pp. 683–685, Feb. 2003. [24] J. K. Park, C. H. Kim, S. H. Park, H. D. Park, and S. Y. Choi, “Application of strontium silicate yellow phosphor for white lightemitting diodes,” Appl. Phys. Lett., vol. 84, pp. 1647–1649, Mar. 2004. [25] J. S. Kim, P. E. Jeon, J. C. Choi, H. L. Park, S. I. Mho, and G. C. Kim, “Warm-white-light emitting diode utilizing a single-phase full-color Ba MgSi O : Eu , Mn phosphor,” Appl. Phys. Lett., vol. 84, pp. 2931–2933, Apr. 2004. [26] J. S. Kim, P. E. Jeon, Y. H. Park, J. C. Choi, H. L. Park, G. C. Kim, and T. W. Kim, “White-light generation through ultravioletemitting diode and white-emitting phosphor,” Appl. Phys. Lett., vol. 85, pp. 3696–3698, Oct. 2004. [27] G. A. Hirata, J. Carver, E. J. Bosze, and J. McKittrick, “Long-UV excited white-emitting phosphors,” Proc. SPIE, vol. 4776, pp. 156–161, Nov. 2002. [28] J. K. Sheu, S. J. Chang, C. H. Kuo, Y. K. Su, L. W. Wu, Y. C. Lin, W. C. Lai, J. M. Tsai, G. C. Chi, and R. K. Wu, “White-light emission from near UV InGaN-GaN LED chip precoated with blue/ green/red phosphors,” IEEE Photon. Technol. Lett., vol. 15, no. 1, pp. 18–20, Jan. 2003. [29] M. Yamada, T. Naitou, K. Izuno, H. Tamaki, Y. Murazaki, M. Kameshima, and T. Mukai, “Red-enhanced white-light-emitting diode using a new red phosphor,” Jpn. J. Appl. Phys., vol. 42, pt. 2, pp. L20–L23, Jan. 2003. [30] R.-J. Xie, N. Hirosaki, K. Sakuma, Y. Yamamoto, and M. SiAlON: a yellow phosphor Mitomo, “Eu -doped Ca for white light-emitting diodes,” Appl. Phys. Lett., vol. 84, pp. 5404–5406, Jun. 2004. [31] A. Žukauskas, F. Ivanauskas, R. Vaicekauskas, M. S. Shur, and R. Gaska, “Optimization of multichip white solid-state lighting source with four or more LEDs,” Proc. SPIE vol. 4445, pp. 148–155, Dec. 2001. [32] A. Žukauskas, R. Vaicekauskas, F. Ivanauskas, R. Gaska, and M. S. Shur, “Optimization of white polychromatic semiconductor lamps,” Appl. Phys. Lett., vol. 80, pp. 234–236, Jan. 2002. [33] A. Žukauskas, R. Vaicekauskas, F. Ivanauskas, M. S. Shur, and R. Gaska, “Optimization of white all-semiconductor lamp for solid-state lighting applications,” Int. J. High Speed Electron. Syst., vol. 12, pp. 429–437, Jun. 2002. [34] W. A. Thornton, “Luminosity and color-rendering capability of white light,” J. Opt. Soc. Amer., vol. 61, pp. 1155–1163, Sep. 1971. [35] A. Zukauskas, R. Vaicekauskas, F. Ivanauskas, G. Kurilcik, Z. Bliznikas, K. Breive, J. Krupic, A. Rupsys, A. Novickovas, P. Vitta, A. Navickas, V. Raskauskas, M. S. Shur, and R. Gaska, “Quadrichromatic white solid-state lamp with digital feedback,” Proc. SPIE vol. 5187, pp. 185–198, Jan. 2004. [36] A. Zukauskas, V. Viliunas, R. Stanikunas, A. Svegzda, H. Vaitkevicius, Z. Bliznikas, K. Breive, R. Vaicekauskas, A. Novickovas, G. Kurilcik, R. Gaska, and M. S. Shur, “Color perception under illumination by quadrichromatic solid-state lamp,” Proc. SPIE, vol. 5530, pp. 347–353, Oct. 2004. [37] F. Ivanauskas, R. Vaicekauskas, A. Žukauskas, R. Gaska, and M. S. Shur, “Stability of the quadrichromatic solid-state lamp,” Proc. Inst. Phys. Conf. Series, vol. 182, pp. 399–400, 2004, in . [38] S. Muthu, F. J. P. Schuurmans, and M. D. Pashley, “Red, green, and blue LEDs for white light illumination,” IEEE J. Sel. Topics Quantum Electron., vol. 8, no. 2, pp. 333–338, Mar./Apr. 2002. [39] D. B. Judd and G. Wyszecki, Color in Business, Science and Industry. New York: Wiley, 1975. [40] G. Wyszecki and W. S. Stiles, Color Science: Concepts and Methods, Quantitative Data and Formulae. New York: Wiley, 2000. [41] J. J. Wierer, D. A. Steigerwald, M. R. Krames, J. J. O’Shea, M. J. Ludowise, G. Christenson, Y.-C. Shen, C. Lowery, P. S. Martin, S. Subramanya, W. Götz, N. F. Gardner, R. S. Kern, and S. A. Stockman, “High-power AlGaInN flip-chip light-emitting diodes,” Appl. Phys. Lett., vol. 78, pp. 3379–3381, May 2001. 0 0 1702 [42] M. R. Krames, M. Ochiai-Holcomb, G. E. Höfler, C. Carter-Coman, E. I. Chen, I.-H. Tan, P. Grillot, N. F. Gardner, H. C. Chui, J.-W. Huang, S. A. Stockman, F. A. Kish, M. G. Craford, T. S. Tan, C. P. Kocot, M. Hueschen, J. Posselt, B. Loh, G. Sasser, and D. Collins, “High-power truncated-in) In P/GaP light-emitting diodes verted-pyramid (Al Ga exhibiting >50% external quantum efficiency,” Appl. Phys. Lett., vol. 75, pp. 2365–2367, Oct. 1999. [43] R. Mueller-Mach, G. O. Mueller, M. R. Krames, and T. Trottier, “High-power phosphor-converted light-emitting diodes based on III-nitrides,” IEEE J. Sel. Topics Quantum Electron., vol. 8, no. 2, pp. 339–345, Mar./Apr. 2002. [44] Organic Light-Emitting Devices J. Shinar Ed. New York: Springer-Verlag, 2004. [45] J. Nylander, “Light on the subject,” Stage Directions, Mar. 2005. [46] J. Tsao, “Roadmap projects significant LED penetration of lighting market by 2010,” Laser Focus World, vol. 39, pp. S11–S14, May 2003. Michael S. Shur (Fellow, IEEE) received the M.S.E.E. degree (with honors) from St. Petersburg Electrotechnical Institute, St. Petersburg, Russia, in 1965, the Ph.D. (candidate) and D.Sc. degrees in physics and mathematics from A. F. Ioffe Institute of Physics and Technology, St. Petersburg, in 1967 and 1992, respectively, and the Honorary Doctorate degree from Saint Petersburg State Technical University, St. Petersburg, in 1994. He has held research or faculty positions at different universities, including A.F. Ioffe Institute, Cornell University, the University of Minnesota, and the University of Virginia, where he was John Money Professor of Electrical Engineering and served as Director of Applied Electrophysics Laboratories. He is now Patricia W. and C. Sheldon Roberts ’48 Professor of Solid State Electronics, Professor of Physics, Applied Physics and Astronomy, and Director of Center for Broadband Data Transport Science and Technology at Rensselaer Polytechnic Institute, Troy, NY. He is Editor-in-Chief of the International Journal of High Speed Electronics and Systems, Editor of a book series on Special Topics in Electronics and Systems published by World Scientific, the Regional Editor of Physica Status Solidi, a Member of the Honorary Board of Solid State Electronics, and a Member of the International Advisory Committee of the Journal of Semiconductor Technology and Science. He has published many books and technical papers and has over 30 patents on electronic devices. Several of his papers received best poster paper awards at MRS and other conferences. He is listed as one of the “most quoted researchers in his field.” He is also one of the co-developers of AIM-Spice, which has over 50 000 users. His current research interests include nanoelectronics and novel THz and photonic devices. Dr. Shur received the van der Ziel Award from ISDRS-99 and Commendation for Excellence in Technical Communications from Laser Focus World in 1999. In 2002, he was awarded the Senior Humboldt Research Prize. In 2003, he received the Pioneer Award from Semiconductor Semi and received the RPI School of Engineering Research Award. He is a Fellow of the American Physical Society, a Fellow of the Electrochemical Society, a Fellow of the World Innovation Foundation, a member of Eta Kappa Nu, Sigma Xi, and Tau Beta Pi, Electromagnetic Academy, Materials Research Society, the American Society for Engineering Education, SPIE, and the American Association for the Advancement of Science, and is an elected member, former Chair, and former elected Member-at-Large of the National Council of U.S. Commission D, International Union of Radio Science. In 1990–1993, he served as an Associate Editor of IEEE TRANSACTIONS ON ELECTRON DEVICES. He has served as Chair, Program Chair, and Organizing and Program Committee Member of many IEEE conferences and as Member and Chair of the IEEE Awards Committee. He is now serving as Vice-President for Publications of the IEEE Sensor Council. In 2003, he was appointed IEEE Electron Device Society Distinguished Lecturer. In 2004, he was appointed IEEE Distinguished Microwave Lecturer. PROCEEDINGS OF THE IEEE, VOL. 93, NO. 10, OCTOBER 2005 Artūras Žukauskas received the Ph.D. (candidate) degree in semiconductor and dielectric physics and the second degree (habilitation doctor) in natural sciences from Vilnius University, Vilnius, Lithuania, in 1983 and 1991, respectively. Since 1979, he has been with Vilnius University; at present he is a Director and Chief Research Scientist at the Institute of Materials Science and Applied Research, and also a Professor at the Department of Semiconductor Physics. In 2000 and 2001, he was also with Rensselaer Polytechnic Institute, Troy, NY, as a Visiting Scientist. He has published a book and over 150 technical papers. Currently, his research group works on the fundamental investigations of carrier relaxation and recombination in highly photoexcited direct-gap semiconductors, optical characterization of materials for optoelectronics (mostly AlGaInN system), and applications of light-emitting diodes in lighting, phototherapy, plant cultivation, optical measurements, and fluorescence sensing of chemical and biological agents. Dr. Žukauskas was awarded the Lithuanian National Prize in Science in 2002. He is a Member of the Lithuanian Physical Society, a Member of the Lithuanian Materials Research Society, and an Expert Member of the Lithuanian Academy of Sciences. He has contributed to numerous Lithuanian national and international conferences as a Chair and Organizing and Program Committee Member. SHUR AND ŽUKAUSKAS: SOLID-STATE LIGHTING: TOWARD SUPERIOR ILLUMINATION 1703