Development of phosphors with high thermal stability and efficiency for phosphor-converted LEDs
© Tian; licensee Springer 2014
Received: 26 February 2014
Accepted: 20 June 2014
Published: 1 August 2014
This article briefly reviews the developments of the phosphors with high thermal stability and high efficiency for use in phosphor-converted LED (pcLED) packages by Lightscape Materials Inc. (wholly owned by The Dow Chemical Company). The current industry design objectives for pcLED packages are first outlined with emphasis on the emission spectral features required for target brightness and color characteristics for general illumination and back light unit for liquid crystal displays. There is a growing demand on thermal stability of the phosphor emission as the LED package power increases. A phenomenological analysis is described on luminescence loss and its relation to thermal stability of luminescence emission at elevated temperatures, which serves as an empirical guide in the search for new phosphor materials of high thermal stability. Finally, the formulations and luminescence properties of the proprietary carbidonitride and oxycarbidonitride phosphors are discussed.
KeywordsPhosphor LED Phosphor-converted LED Carbidonitide Oxycarbidonitride
The phosphor-converted light emitting diode (pcLED) combines a GaN-based LED with down-converting phosphors and emits light in the visible spectral region. The phosphor beneficially absorbs a portion of the blue or near UV (nUV) LED excitation emission of the native GaN-based LED and broadens the emission spectrum by re-emitting at longer wavelengths. Since high brightness (HB) GaN-based LED became available in the mid-1990’s, the pcLED has been rapidly developing into many applications, ranging from back light units (BLU) for liquid crystal displays (LCD), to general illumination, and to many specialty and niche lighting areas ,. As pcLED products continue to grow in diversity of package designs, their luminescent features and performance are largely determined by the phosphors used in the packages. The development and availability of the phosphor, therefore, will be a decisive factor to the eventual penetration of pcLEDs into the marketplace.
It had been an industry wish to produce light with LEDs by converting the LED emission into an extended wavelength range in the visible light spectrum ever since semiconductor-based LEDs were developed. The concept of using phosphors to convert the LED emission was demonstrated by an LED chip packaged with a phosphor –. The earliest demonstrations started in the 1970’s, including, for example, a SiC-based, yellow-emitting LED chip packaged with a rhodamine-containing resin to generate orange-red light through down-conversion, and a GaAs-based infrared LED chip packaged with LaF3:Tb,Er phosphor to generate green light through up-conversion. Although the invention of GaN-based, blue-emitting LEDs brought about the promise of providing a red-green-blue (RGB) light source for full color displays and a white light source for general illumination in the early 1970’s, it was not possible to realize that promise until the GaN-based LEDs achieved high brightness performance in mid-1990’s.
The emergence of HB GaN-based LEDs triggered an intense search for phosphors that are efficiently excited by blue or nUV light. The initial search included a comprehensive screening of the traditional phosphors discovered and investigated over some 60 years mainly for cathode-ray tube (CRT) and fluorescent lamps. These activities are reflected in the related patent literature – and summarized in a chapter in a recent book . The search was clearly motivated by, and focused on, white light generation to meet a long-standing industrial goal. As a result, yttrium aluminum garnet doped with cerium (YAG:Ce) was identified as a high efficiency phosphor whose broad band, yellow emission combined with a blue LED emission produces a good white light ,. In parallel with the success of YAG:Ce, alkaline earth orthosilicates doped with europium (BOSE)  were quickly recognized as an efficient yellow emitter, and adopted in low power pcLED packages (<1Watt/chip) as an alternative for YAG:Ce.
On the other hand, the development of both GaN-based LED chips and pcLED packaging technique continued to move toward increasingly higher power pcLED packages (>1 Watt/chip), driven by the needs for the BLU for large screen TVs and general illumination. However, a unique aspect of conventional pcLED is that the phosphors used in such packages are in physical contact with the LED chip, and the LED chips operate at high temperatures (e.g., in the range of 100°C-150°C). Traditional phosphors (e.g., BOSE) exhibit a significant drop in efficiency at such operating temperatures, limiting the performance of the pcLED devices. This stimulated another round of discovery for phosphor formulations which were required to operate efficiently at high temperatures in high power LED packages. The primary criterion for the phosphor is high thermal stability of luminescence with a lumen maintenance at 150°C of >90% that at room temperature. Naturally, the search for good host crystals started among refractory materials such as nitrides with a high melting point. As a result, several nitride and oxynitride phosphors were successfully formulated in the early 2000’s, and in turn applied to pcLED products. The formulations include the red-emitting CaAlSiN3:Eu , and Sr2Si5N8:Eu ,, the yellow-emitting Ca-α-sialon:Eu , and La3Si6N11:Ce , the green-emitting β-sialon:Ce  and SrSi2N2O2:Eu . All these phosphors are converters of blue or nUV light and have high thermal stability. The application of these phosphors not only enabled the pcLED products operating at > 1 Watt per single chip package, but also significantly expanded the luminescence features of pcLED packages.
There has been an extensive investigation on the nitride and oxynitride phosphors in the last decade. Several review articles have been published that summarize the accomplishments in the field before the year 2011 ,–. The host crystal structures of these phosphors are based on a network of tetrahedral [SiN4], [Si(O,N)4], [AlN4] or [Al(O,N)4] which are interconnected to each other through corner-sharing nitrogen atoms. Unlike in most traditional phosphors where the luminescence activators bond to halides or oxygen, the activators, Eu2+ or Ce3+, are coordinated directly with nitrogen atoms. The coordinating bonds thus are more covalent and have higher polarizability than those with halides and oxygen (i.e., Nephelauxetic effect) . This coordinating environment results in a relatively large centroid shift of 4f-5d transition energy which corresponds to an optical absorption and emission in relatively low energy. Typically, the absorption of Eu2 + −activated nitride phosphors is in blue to nUV spectral region while the emission is in green to red spectral region ,–. Inspired by the research results, many different crystalline nitrides and oxynitrides have been studied as candidate phosphor host crystals. This article is intended to give a brief overview of the developments of proprietary phosphors for use in pcLED packages by Lightscape Materials Inc. (wholly owned by The Dow Chemical Company). The current industry design targets to be incorporated into pcLED packages will be reviewed with an emphasis on the emission features desired for target brightness and color quality. A phenomenological analysis will be described on luminescence loss at high temperatures, which is related to thermal stability of luminescence emission. Finally, proprietary carbidonitride and oxycarbidonitride phosphor formulations will be discussed.
Requirements for LED Phosphors
Desired luminescence features
High quantum efficiency
Strong absorption in blue and nUV spectral range
High thermal stability of luminescence
Short emission decay time
Long term stability
Low cost and low materials usage
Environmentally benign composition
Requirement 1, desired luminescence features, includes the spectral features of luminescence excitation and emission. This requirement often serves as the first gate for screening. The requirements 2 to 4 are the intrinsic properties that determine the efficiency performance of the phosphors. While these properties could usually be improved by better crystallization and purifying materials, a threshold value is usually set for each of them at the early stage of development, e.g., quantum efficiency > 60%. Requirement 5, short emission decay time, is related to the emission saturation of the phosphor when it is excited with high photon flux. If a phosphor material with a long decay time is excited with a high photon flux, the emission efficiency can be lower than the value at low power excitation . Requirement 6, long term stability, directly impacts on the life time of the pcLED products which is related to the chemical stability and degradation of the phosphor materials under operating conditions. Requirement 7, the cost and low materials usage, is increasingly important as further cost reduction of pcLEDs enables further implementation of solid state lighting. From the phosphor developer’s perspective, this requirement can be met by improving in two directions: (1) reduction of the cost of raw materials and processing and (2) reduction of the mass of the phosphor required in pcLED packages to achieve the designed color specifications. Requirement 8, environmentally benign composition, limits the selection of constituent elements and hence the chemical formulations of the phosphors to be devised. Certain elements such as cadmium do not meet this requirement and formulations containing such elements should be ruled out.
Currently available phosphors have enabled pcLED products to deliver lumen performance at the level of fluorescent lamps. However, there is still a significant gap between today’s best-in-class product performance and the target performance for fully realizing the potential benefits of LED lighting. For example, most general lighting products have efficiency below 100 lm/W, while the DOE target for 2020 is 200 lm/W to achieve the 19% energy savings in lighting relative to the year 2010 ,. In addition to the main goal of energy savings, the pcLED product must demonstrate a compelling value in the marketplace before it begins to win sizeable market share from the incumbent products. The value needs to be reflected in the color performance, light output, efficacy, reliability, cost, lifetime, and manufacturability, many of which depend on to what degree the phosphors meet the above seven requirements.
The big gap in LER indicates that there is a compelling need for improvement in phosphor emission features and efficiency performance. One of the sources of lumen deficiency is the portion of the broad red emission band whose radiance quanta are outside of the luminous efficiency curve. Narrowing the bandwidth of the red emission while maintaining its redness is desired. A simulation with a red LED (FWHM ~ 25 nm) combined with a blue LED + YAG:Ce demonstrated a white light of LER = 375 lm/W at CCT = 3000 K and CRI = 90 –. This prompts an approach to use a red line-emitting phosphor in place of the red LED. Recently, a red line-emitting, Mn4+-activated phosphor (K2SiF6:Mn4+, i.e., KSF:Mn4+) has been shown to produce a white light at CCT = 3510 K and CRI = 90.9 in a package with a blue LED and YAG:Ce phosphor ,. Compared to the baseline white pcLED with a blue LED and YAG:Ce phosphor, the lumen performance of the KSF:Mn4+-assisted white pcLED is still low. This demonstration shows promise to achieve a high quality white light with a line-emitting red phosphor packaged in a pcLED.
The potential lumen performance of line-emitting phosphors in a white light source can be understood from the experience with fluorescent lamps. In the tri-phosphor luminescent lamp, a combination of three line emitting phosphors was well known to produce white light with high LER and high color quality. The three line emitting phosphors, e.g., in Color 80 Lamps, are BaMgAl10O17:Eu2+, (Ce,Gd)MgB5O10:Tb3+ and Y2O3:Eu3+, which emit lines at ~450 nm, ~540 nm and ~610 nm, respectively, and they together generate a white light of CRI 80–85 . LER and CRI of such a white light depend critically on the exact wavelength position of the emission lines of the blue and red phosphor whereas they are less sensitive to the wavelength position of the green emission line. To produce a white light with CRI > 80, the blue emission line is required to be in the range of 440–490 nm and the red emission line in 595–618 nm –. Learning from the above facts, it is considered a possible path to improving LER with line-emitting phosphors in pcLEDs. With GaN-based LED as the narrow band blue emitter, line emitting green and red phosphors would form a tri-color white light pcLED with high LER and high color quality. Several red line emitting phosphors have been formulated. These include K2SiF6:Mn4+, K2TiF6:Mn4+, Na2SnF6:Mn4+, Cs2SnF6:Mn4+, Na2SiF6:Mn4+, Na2GeF6:Mn4+, and β-sialon:Pr3+. These phosphors can be excited with blue or nUV light and emit a line spectrum of red color.
For BLU applications, the conventional pcLED packages offer ~72% of NTSC ratios while next generation products aim at an NTSC ratio of close to 100%. Moreover, pcLED packages for BLU need to be dimmable over a wider power range while also exhibiting an excellent long term stability. These requirements are related to the emission features and performance of the phosphor, some of which necessitate the development of innovate new phosphor formulations.
Thermal stability of luminescence emission
Traditional phosphors in CRTs and fluorescent lamps are excited by a relatively low energy flux density. In these operating systems, a relatively large amount of phosphor powder absorbs the excitation energy and is positioned at a distance from the excitation source. In contrast, the phosphors in pcLED packages are excited by a significantly higher energy flux density, with much less phosphor powder placed in direct contact with the LED chip. Under these operating conditions, the phosphors not only take up thermal energy directly from the chip (typical junction temperature ~100°C) but also are excited by a high density of photon energy from the chip. Being heated, the crystal of the phosphors is at a high vibrationally excited state, causing the LED excitation energy to be directed more to heat emission through lattice relaxation rather than to the luminescence emission. This relaxation process corresponds to Stokes shift which produces further heating, thereby further reducing the luminescence emission. This is a vicious cycle that precludes applications of the traditional phosphors in pcLEDs. Successful development of high power pcLEDs for general illumination, therefore, requires the development of phosphors that can emit highly efficiently at temperatures of 100°C-150°C (Requirement 4 as mentioned above). It is critical that we develop phosphor materials with low thermal quenching at temperatures of 100–200°C.
Hardness and thermal expansion values
Thermal expansion, 10−6/K
The luminescence features of the phosphors F540, F550 and F560
Emission Bandwidth (FWHM)
CIE 1931 Chromaticity Coordinates
x = 0.347, y = 0.615
x = 0.383, y = 0.590
x = 0.436, y = 0.550
The investigation of high performance phosphor formulations for pcLED has been driven by the industrial needs for general illumination and BLU for LCD displays. After the initial effort to identify competitive phosphors among the traditional formulations, e.g., YAG:Ce3+ and MSiO4:Eu2+, eight criteria were identified for the qualification of phosphor formulations. An empirical analysis on the energetics of photophysical events was conducted based on the energy curves in configurational coordinates that provides guidance in the search for next generation phosphor formulations, i.e., the high hardness of the host crystalline materials is correlated to high thermal stability of the luminescence emission. Proprietary phosphor formulations of carbidonitride and oxycarbidonitride have been investigated as phosphors with high efficiency and high thermal stability for use in pcLED devices.
The author gratefully appreciates Lora Brehm, Nate Brese, Towhid Hasan, Anne Leugers, Yuan Qiang Li, Mark McAdon, Michael Romanelli, Robert Simms, Alan Thomas and Britt Vanchura at The Dow Chemical Company, whose competence and dedication have created the knowledge outlined in this article. The author also thanks one of the anonymous reviewers for the many valuable suggestions and comments.
- Solid State Lighting Research and Development Multi-Year Program Plan. In April 2013, Prepared for Lighting Research and Development, Office of Energy Efficiency and Renewable Energy U.S. Department of Energy. ., [http://www1.eere.energy.gov/buildings/ssl/]
- Azevedo IL, Morgan MG, Morgan F: The transition to solid-state lighting. Proc IEEE 2009, 97(3):481–510.View ArticleGoogle Scholar
- Geusic JE, Ostermayer FW, Marcos HM, Van Uitert LG, van der Ziel JP: J Appl Phy 1971, 42(5):1958.View ArticleGoogle Scholar
- Porret F: US Patent 4,038,580.
- Potter RM: US Patent 3,529,200.
- Amans RL: US Patent 3,510,732.
- O'Connor JM, Aina OA: US Patent 5,208,462.
- Moyer CD, Jaskie JE, Legge RN: US Patent 5,334,855.
- LED phosphors and down-converters patent investigation. In Yole Developpement. 2013.
- Baretz B, Tischler M: US Patent 6,600,175.
- Reeh U, Höhn K, Stath N, Waitl G, Schlotter P, Schneider J, Schmidt R: US Patent 6,576,930.
- Xie R-J, Li YQ, Hirosaki N, Yamamoto H: Nitride Phosphors and Solid-State Lighting. CRC Press, Boca Raton; 2011.Google Scholar
- Collins WD III, Krames MR, Verhoeckx GJ, van Leth NJ M: US Patent 6,642,652.
- Shimizu Y, Sakano K, Noguchi Y, Moriguchi T: US Patent 5,998,925.
- Srivastava A, Comanzo HA, McNulty TF: US Patent 6,621,211.
- Uheda K, Hirosaki N, Yamamoto H: Host lattice materials in the system Ca3N2–AlN–Si3N4for white light emitting diode. Phys Status Solidi a 2006, 203: 2712.View ArticleGoogle Scholar
- Uheda K, Hirosaki N, Yamamoto Y, Naito A, Nakajima T, Yamamoto H: Electrochem Solid-State Lett 2006, 9: H22.View ArticleGoogle Scholar
- Hoppe HA, Lutz H, Morys P, Schnick W, Seilmeier A: Luminescence in Eu2+-doped Ba2Si5N8: fluorescence, thermoluminescence, and upconversion. J Phys Chem Solids 2000, 61: 2001.Google Scholar
- Li YQ, van Steen JEJ, van Krevel JWH, Botty G, Delsing ACA, DiSalvo FJ, de With G, Hintzen HT: Luminescence properties of red-emitting M2Si5N8:Eu2+ (M = Ca, Sr, Ba) LED conversion phosphors. J Alloys Compounds 2006, 417: 273–279.View ArticleGoogle Scholar
- van Krevel JWH, van Rutten JWT, Mandal H, Hintzen HT, Metselaar R: Luminescence properties of terbium-, cerium-, or europium-doped α -sialon materials. J Solid State Chem 2002, 165: 19–24.View ArticleGoogle Scholar
- Xie RJ, Mitomo M, Uheda K, Xu FF, Akimune Y: Preparation and luminescence spectra of calcium- and rare-earth (R = Eu, Tb, and Pr)-codoped α-SiAlON ceramics. J Am Ceram Soc. 2002, 85: 1229–1234.View ArticleGoogle Scholar
- Kijima N, Seto T, Hirosaki N: A new yellow phosphor La3Si6N11:Ce3+for white LEDs. In 216th ECS Meeting, Electrochem Soc 2009, 25: 247.View ArticleGoogle Scholar
- Hirosaki N, Xie RJ, Kimoto K, Sekiguchi T, Yamamoto Y, Suehiro T, Mitomo M: Characterization and properties of green-emitting β-SiAlON:Eu2+powder phosphors for white light-emitting diodes. Appl Phys Lett 2005, 86: 211905.View ArticleGoogle Scholar
- Li YQ, Delsing ACA, de With G, Hintzen HT: Luminescence properties of Eu2+-activated alkaline-earth silicon-oxynitride MSi2O2-δN2+2/3δ (M = Ca, Sr, Ba): a promising class of novel LED conversion phosphors. Chem Mater 2005, 17: 3242–3248.View ArticleGoogle Scholar
- Xie RJ, Hirosaki N: Silicon-based oxynitride and nitride phosphors for white LEDs—A review. Sci Tech Adv Mater 2007, 8: 588–600.View ArticleGoogle Scholar
- Xie R-J, Hirosaki N, Li Y, Takeda T: Rare-earth activated nitride phosphors: synthesis, luminescence and applications. Materials 2010, 3: 3777–3793.View ArticleGoogle Scholar
- Zeuner M, Pagano S, Schnick W: Nitridosilicates and oxonitridosilicates: from ceramic materials to structural and functional diversity. Angew Chem Int Ed 2011, 50: 7754–7775.View ArticleGoogle Scholar
- Dorenbos P: Ce3+ 5d-centroid shift and vacuum referred 4f-electron binding energies of all lanthanide impurities in 150 different compounds. J. Lumin 2013, 135: 93–104.View ArticleGoogle Scholar
- Smet PF, PArmentier AB, Peolman D: Selecting conversion phosphors for white light-emitting diodes. J Electrochem Soc 2011, 158: R37-R54.View ArticleGoogle Scholar
- Ohno Y: Improving the color spectrum to increase LED efficacy, In DOE Solid State Lighting R&D Workshop, February 2–4: Raleigh. North: Carolina; 2010.Google Scholar
- Ohno Y: Proc. SPIE 5530, 4th Int. Conf. on Solid State Lighting, Denver, CO 2004.Google Scholar
- Ohno Y: Optical Engineering 2005, 44(11):30.View ArticleGoogle Scholar
- Ohno Y, Davis W: Proc. SPIE 5941, 5th Int. Conf. on Solid State Lighting, San Diego, CA 2005.Google Scholar
- Setlur AA, Radkov EV, Henderson CS, Her JH, Srivastava AM, Karkada N, Kishore MS, Kumar NP, Aesram D, Deshpande A, Kolodin B, Grigorov LS, Happek U: Energy-efficient, high-color-rendering LED lamps using oxyfluoride and fluoride phosphors. Chem Mater 2010, 22: 4076–4082.View ArticleGoogle Scholar
- Liao C, Cao R, Ma Z, Li Y, Dong G, Kaniyarakkal N, Sharafudeen , Qiu J: Synthesis of K2SiF6:Mn4+ phosphor from SiO2 powders via redox reaction in HF/KMnO4 solution and their application in warm-white LED. J Am Ceram Soc 2013, 96: 3552–3556.View ArticleGoogle Scholar
- Jüstel T, Nikol H, Ronda C: New developments in the field of luminescent materials for lighting and displays. Angew Chem Int Ed 1998, 37: 3084–3103.View ArticleGoogle Scholar
- Thornton WA: Luminosity and color-rendering capability of white light. J Opt Soc Am 1971, 61: 1155–1163.View ArticleGoogle Scholar
- Koedam M, Opstelten JJ: Light. Res. Technol. 1971, 3: 205.Google Scholar
- Verstegen JMPJ, Radielovic D, Vrenken LE: A new generation of “deluxe” fluorescent lamp, combinging an efficiency of 80 lumens/W or more with a color rendering index of approximately 85. J Electrochem Soc 1974, 121: 1627–1631.View ArticleGoogle Scholar
- Arai Y, Adachi S: Optical properties of Mn4+-activated Na2SnF6 and Cs2SnF6 red phosphors. J Lumin 2011, 131: 2652–2660.View ArticleGoogle Scholar
- Xu YK, Adachi S: Properties of Na2SiF6:Mn4+and Na2GeF6:Mn4+red phosphors synthesized by wet chemical etching. J Appl Phys 2009, 105: 013525.View ArticleGoogle Scholar
- Liu TC, Cheng BM, Hu SF, Liu RS: Highly Stable Red Oxynitride β-SiAlON:Pr3+ Phosphor for Light-Emitting Diodes. Chem Mater 2011, 23: 3698–3705.View ArticleGoogle Scholar
- Leveranz HW: An Introduction to Luminesce of Solids. Dover Publications, New York; 1950.Google Scholar
- Li Y, Romanelli M, Tian Y: Carbidonitride- and oxycarbidonitride-based phosphors for LED lighting devices. In Proceeding SPIE: Gallium Nitride Materials and Devices VII 2012, 8262: 82621O-1.View ArticleGoogle Scholar
- Tian Y: Nitride and Oxynitride Based Phosphors for Solid State Lighting. In: DOE SSL R&D Workshop, San Diego, CA, February 2, 2011.View ArticleGoogle Scholar
- Bachmann V, Ronda C, Meijerink A: Temperature quenching of yellow Ce3+ luminescence in YAG:Ce. Chem. Mater 2009, 21: 2077–2084.View ArticleGoogle Scholar
- Brgoch J, DenBaars SP, Seshadri R: Proxies from Ab initio calculations for screening efficient Ce3+phosphor hosts. J Phys Chem C 2013, 117: 17955.View ArticleGoogle Scholar
- Li Y, Romanelli MD, Tian Y: Silicon carbidonitride-based phosphors and lighting devices using the same. US Patent 8,535,566.
- Li Y, Romanelli MD, Tian Y: Carbidonitride phosphors in LED lighting products. Global phosphor summit, March 19, 2013; New Orleans, LA.Google Scholar
- Nevskii NN, Glasser LD, Iliukhin VV, Belov NV: Sov. Phys. Crystallogr. 1979, 24: 93.Google Scholar
- Li Y, Romanelli MD, Tian Y: Oxycarbidonitride phosphors and light emitting devices using the same. US Patent 8,551,361.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.