New broadband glass phosphors with excellent thermal stability were proposed and experimentally demonstrated for white light-emitting-diodes (WLEDs). The novel glass phosphors were realized through dispersing multiple phosphors into SiO2 based glass (SiO2-Na2O-Al2O3-CaO) at 680°C. Y3Al5O12:Ce3+ (YAG), Lu3Al5O12:Ce3+ (LuAG), and CaAlSiN3: Eu2+ (nitride) phosphor crystals were chosen respectively as the yellow, green, and red emitters of the glass phosphors. The effect of sintering temperature on inter-diffusion reduction between phosphor crystals and amorphous SiO2 in nitride-doped glass phosphors was studied and evidenced by the aid of high-resolution transmission electron microscopy (HRTEM). Broadband glass phosphors with high quantum-yield of 55.6% were thus successfully realized through the implementation of low sintering temperature. Proof-of-concept devices utilizing the novel broadband phosphors were developed to generate high-quality cool-white light with trisstimulus coordinates (x, y) = (0.358, 0.288), color-rending index (CRI) = 85, and correlated color temperature (CCT) = 3923K. The novel broadband glass phosphors with excellent thermal stability are essentially beneficial to the applications for next-generation solid-state indoor lighting, especially in the area where high power and absolute reliability are required.
© 2014 Optical Society of America
Color rendering index (CRI), which indicates how light sources accurately render the colors of objects , has become a key metric for high-quality illumination. To ensure objects appear natural and vivid, a CRI criterion of 80 for indoor lightings has been specified by U.S. ENERGY STAR . Conventional light sources such as incandescent light bulbs and fluorescent lamps typically exhibit high CRIs. However, they are considerably energy consuming. In the aspect of energy saving, light-emitting diodes (LEDs) have been vigorously developed as alternatives to conventional light sources [3–6]. For realization of white light, two or more main wavelengths are required. Considering both cost and complexity of driving circuits, the most common way to produce additional colors in LEDs is to use color conversion materials that absorb light emitted from the LEDs and convert the absorbed wavelength to longer wavelengths. Y3Al5O12:Ce3+, a phosphor with peak absorption at 458nm and a peak emission at 560nm, is widely used as the color conversion material for WLEDs . However, the CRI of the WLEDs that employ only single phosphor is usually less than 70, which is not satisfactory for the applications in high-quality lighting. To achieve high color rendering properties, color conversion layers with red emission in addition to yellow emission are essential [8, 9] for broad wavelength coverage. Such broadband color conversion layers have been developed by doping multiple phosphors into silicone-based matrix . The CRI of the WLEDs utilizing the multi-phosphor-doped silicone (MPDS) is above 90, whereas the poor thermal stability of the silicone-based color conversion layers weakens the superiority for lighting applications [11, 12].
In recent years, glass-based phosphors have been discussed as the solution of conversion layers for high-power WLEDs due to much higher thermal stability than silicone-based phosphors. Glass-based phosphors can be realized by using many kinds of methods including re-crystallization and sol-gel [13, 14], but high processing temperature raises the fabrication cost of these methods. Dispersing phosphor crystals into glass matrix directly is thus an alternative way to realize glass-based phosphors, because its processing temperature is determined by the melting point of the glass, which can be considerably lower. Lead, tellurite, and borate glasses have been used as the matrices of phosphor crystals [15–18]. The lead and tellurite glass phosphors can be successfully sintered at 700°C and 500°C, respectively. Lead glass phosphors show high refractive index, but the containing lead element may cause environmental pollution issues. Tellurite glass phosphors exhibit the lowest sintering temperature in the category of glassed-based phosphors, but their pale yellow appearance and strong crystallization tendency limit their optical performance. Borate glass phosphors present high optical transparency, but their sintering temperature is up to 750°C, which is much higher than other glass phosphors. In our previous work, we have demonstrated thermal-stable sodium glass phosphors for WLEDs [19–21]. The sintering temperature of the sodium glass phosphors is low (700°C), and the optical transparency of the sodium glass phosphors is also good. Owing to this achievement, we believe that multi-phosphor-doped glass (MPDG) would be a good candidate for the applications in high-quality lightings. However there has no study about high-CRI WLEDs utilizing glass-based phosphors yet. The fabricating temperature of such phosphor-doped glass is usually much higher than that of phosphor-doped silicone, so it could be the answer of why the high CRI WLEDs have not been demonstrated in the system of glass-ceramic phosphors and glass phosphors. In this work, optical properties and microstructure of single-phosphor-doped glass (SPDG) sintered at different temperature were investigated to disclose the effect of sintering temperature on performance of glass phosphors. Based on these results, we proposed a newly developed broadband glass phosphors to realize high CRI WLEDs. The thermal stability of the broadband glass phosphors and the optical performance of the WLEDs utilizing the broadband glass phosphors were also demonstrated in this paper.
2. Experimental methods
2.1 Fabrication of glass phosphors
Two key steps were involved in fabricating glass phosphors: i) preparation of sodium mother glass; ii) dispersing Ce3+:YAG phosphor crystals into glass matrix. The composition of the sodium mother glass was 60 mol% SiO2, 25 mol% Na2CO3, 9 mol% Al2O3, and 6 mol% CaO. The raw materials were uniformly mixed and melted at 1300°C, followed by a gradual cooling to room temperature. The resultant cullet glass (SiO2-Na2O-Al2O3-CaO) was dried and milled into powders. Figure 1 illustrates the sequence of fabricating glass phosphors after finishing the preparation of sodium mother glass. Glass phosphor precursor was the uniform mixture of phosphor crystals and the glass powders. Y3Al5O12:Ce3+ (YAG based), Lu3Al5O12:Ce3+ (LuAG based), and CaAlSiN3: Eu2+ (nitride based) phosphor crystals were chosen as the yellow, green and red color conversion elements in glass phosphors, respectively. To disperse the phosphor crystals into the mother glass, sintering the precursor at high temperature was necessary. To understand the effect of the sintering temperature on optical performance of glass phosphors, different sintering temperatures (680°C, 700°C, and 750°C) were implemented in different batches, and the resultant glass phosphors were named according to both the color of containing phosphor and the sintering temperature, i.e. YDG-680 indicates the yellow phosphor-doped glass sintered at 680°C, while GDG-700 and RDG-700 respectively indicate the green phosphor-doped glass and red phosphor-doped glass sintered at 700°C. The broadband glass phosphors in this work, named as YGRDG, were realized by multiple-phosphor-doped glass (MPDG) sintered at 680°C. The three phosphors (yellow, green, and red) in the YGRDG were equal in weight. YAG:Ce3+-doped, LuAG:Ce3+-doped, and CaAlSiN:Eu2+-doped silicone (YDS, GDS, and RDS) were also prepared with the same size and shape of the glass phosphors for a comparison. The baking and curing temperature of those silicone phosphors were 150°C, which is usually used commercially.
2.2 Characterization of glass phosphors
Differential thermal analysis (DTA) and thermal shrinkage of glass phosphor precursor were implemented with a Seiko Tg/DTA 7300 at a rate of 5°Cmin−1. The crystallographic phase of glass phosphors was determined with a Bruker D8 X-ray diffractometer. The luminescent spectra of glass phosphors, GaN-based blue LED, and WLED modules employing different glass phosphors, were measured by an integrating sphere equipped with an optical fiber and a CCD detector. Internal quantum yield (QYint), one of essential parameters used as a criterion of color conversion materials in WLEDs, was defined as the ratio of the number of photons emitted from color conversion materials to the number of photons absorbed from the emission of light sources. The number of photons in each wavelength was derived by dividing spectrum distribution by photon energy . The microstructure of the glass phosphors sintered in different temperature was examined by a high-resolution transmission electron microscope (HRTEM, FEI E.O Tecnai F20 G2 MAT S-TWI) equipped with a LaB6 electron gun operating at 200 kV. The thickness of the glass phosphors for HRTEM study was approximately 60 nm, which was created by focused ion beam technique (FIB, SEIKO SMI3050). The glass phosphors were coated with platinum (about 90nm) to prevent charge accumulation during the examination.
2.3 Thermal aging tests
The thermal stability of both YGRDG and MPDS was studied through thermal aging tests. All the phosphors for the thermal aging tests were shaped to entirely cover a blue-LED (λmax = 445nm) and the CIE chromaticity of the resulting WLEDs was at (x, y) = (0.358 ± 0.005, 0.288 ± 0.005) before the test. Batches of phosphors were then aged for 1008 hours separately at the temperature of 150°C, 250°C, 350°C, and 450°C. The internal quantum yield (QYint) of phosphors and the optical performance of the WLEDs utilizing the phosphors, including CRI and CIE chromaticity, were characterized before and after the thermal aging tests to evaluate the thermal stability of phosphors.
2.4 Fabrication of WLEDs
A GaN-based LED module (λmax = 445nm), center-mounted in a reflective cup, was used as the blue light source of WLEDs. The broadband glass phosphors were shaped to a diameter of 15mm and a thickness of 0.5mm to entirely cover the reflective cup.
3. Results and discussion
3.1 Photophysical properties of SPDGs
Figure 2 shows the images and the QYint of YDGs, GDGs, and RDGs sintered at different temperature. The QYint of YDG-680, GDG-680, and RDG-680 are 68%, 67%, and 46%, respectively. As the sintering temperature increases, both the QYint and the transparency of the RDGs (nitride-based glass phosphors) decrease severely, while those of both YDGs and GDGs are hardly changed. Since glass phosphors were sintered under high temperature, degrading optical performance of the phosphor crystals caused by the high processing temperature might be expected. To evaluate the effect of processing temperature on the phosphor crystals, the QYint of YDS, GDS, and RDS were also measured. The QYint of YDS, GDS, and RDS are 69%, 68% and 51%, respectively. The differences of QYint between the glass and silicone phosphors are considerably lower and therefore 680°C is an adequate fabrication temperature for our glass phosphors.
In the previous paragraph, the QYint of RDGs decays severely with sintering temperature. To further understand the effect of sintering temperature on RDGs, RDGs sintered at different temperature were examined by DTA. As shown in Fig. 3, the RDG precursor shows clear glass transition temperature (Tg), melting temperature (Tm), and crystallization temperature (Tx) at 538°C, 650°C, and 696°C, respectively. The DTA result indicates RDG precursor may be restructured when sintered at the temperature higher than 696°C. To check the crystalline phase of the RDGs sintered in different temperature, the RDGs were examined by X-ray diffraction (XRD). Figure 4 shows the XRD patterns of the RDGs, which confirms the presence of CaAlSiN3: Eu2+ in the RDGs. However, the peak intensity of CaAlSiN3: Eu2 degrades clearly with increasing of sintering temperature, which indicates high sintering temperature indeed weakens the crystalline phase of CaAlSiN3: Eu2+ in glass phosphors. Figure 5 shows HRTEM images of the RDGs, which further confirm the micro-structures in RDGs. A clear boundary only exists between SiO2 and CaAlSiN3: Eu2+ in RDG-680, while a fuzzy region appears between SiO2 and CaAlSiN3: Eu2+ both in RDG-700 and RDG-750. The fuzzy region caused by inter-diffusion between SiO2 and CaAlSiN3: Eu2+ even increases with sintering temperature. The enhanced inter-diffusion between SiO2 and phosphor crystals in considerably high sintering temperature has been discovered in the case of YDGs, and the intrinsic properties of the phosphor crystals have been found to be hindered by the inter-diffusion-induced defects . Through the result of HRTEM, the inter-diffusion in RDGs is found to be much more intense than the inter-diffusion in YDGs. Therefore, ultra-low sintering temperature is critically essential to develop broadband glass phosphors for high-CRI WLEDs.
3.2 Optical properties of the broadband glass phosphor and the resulting WLED
The emission spectra of the broadband glass phosphor, YGRDG, are shown in Fig. 6, and the optical data of both the YGRDG and the resulting WLEDs are listed in Table 1. The data of YDG is also shown as comparison. The peak wavelength and the full width at half maximum of YGRDG are 599 nm and 123 nm, respectively. Figure 6 clearly shows that the emission spectrum of the YGRDG, with the contribution of red and green phosphors, can fill the void in the long wavelength of the conventional YDG and keep good QYint (55.6%) as well.
Color rendering index (CRI, Ra) is one of the most universal metrics for color rendering evaluation. Ra is obtained by the following equation 
where ∆Ei is the color difference (in the 1964 W*U*V* uniform color space) of 8 selected Munsell samples when illuminated by the WLEDs utilizing glass phosphors and when illuminated by a reference illuminant.
By using YGRDG, the resulting WLED provides a CRI of 85 at cool white CCTs, 3923K. The color rendering capability of the WLED utilizing YGRDG is indeed much better than the WLED employing conventional YDG.
3.3 Thermal stability of the broadband glass phosphors
Table 2 shows the results of both YGRDGs and MPDSs under accelerated thermal aging tests. The QYint loss of the phosphors was calculated by the following equation.
where QYi is the quantum yield of the phosphors before aging and QYf is the quantum yield of the phosphors after aging. The QYint loss of YGRDGs aged at 150°C, 250°C, 350°C, and 450°Cis 1.2%, 1.7%, 2%, and 2.2%, respectively. The counterparts, MPDSs, show much poorer thermal stability. The QYint loss of MPDSs is 3.7 and 17.9 times higher than YGRDGs at 150°C and 250°C, respectively. CRI attenuation was calculated by the following equation.
where Rai is the color rendering index of the WLEDs employing non-aged phosphors and Raf is the color rendering index of the WLEDs employing aged phosphors. CRI attenuation of MPDSs is 1.8 and 16.6 after 1008 hours of test at 150°C and 250°C, respectively. However, CRI attenuation is undetectable in the case of YGRDGs, suggesting high reliability of YGRDGs. CIE chromaticity shift caused by thermal test, (ΔE), was derived from the optical data of the WLEDs utilizing the phosphors before and after the thermal aging . As shown in the table, the chromaticity shift of MPDSs is tens of times larger than YGRDGs. Also, the chromaticity shift in YGRDGs is quite stable even under higher temperature. In the case of MPDSs, high heat flux radiation from GaN-base LED chip detaches the methyl group from Si-O frame of silicone, which creates sub-band defects to yellow the silicone and decreases the transmittance of MPDSs. For YGRDGs, the carrier of the fluorescent phosphor crystals is glass, which exhibits glass transition temperature up to 568°C . The glass transition temperature of YGRDGs is high enough to resist thermal stressing. Therefore, YGRDGs show a much better thermal stability than MPDSs in all the aspects we discussed.
In summary, novel broadband glass phosphors with high QYint (55.6%) have been successfully developed. The high QYint of the broadband glass phosphors can be contributed to the low sintering temperature (680°C), which efficiently restrains the inter-diffusion between phosphor crystals and SiO2 evidenced by HRTEM. The WLEDs utilizing the broadband glass phosphors provide high-CRI (85) cool-white light (CCT = 3923K). Cool-white light emitting diodes with high color rendering indices are vitally required for the widespread use of solid state lighting especially indoors, which indicates huge commercial potential of the broadband glass phosphors. The broadband glass phosphors also show an excellent thermal stability, including remarkably low QYint loss, undetectable CRI attenuation, and considerably small chromaticity shift after thermal stressing. This study, which utilizes novel broadband glass phosphors as the color conversion layers in WLEDs, can lead to a creation of high-quality and high-power solid-state lightings.
This work was supported by the National Science Council under the Grants NSC 100-3113-E-110-003-CC2 and the Advanced Optoelectronic Technology Center (AOTC), National Cheng Kung University. Also thanks to professor Jau-Sheng Wang provided glass material.
References and links
1. S. Chhajed, Y. Xi, Y. L. Li, Th. Gessmann, and E. F. Schubert, “Influence of junction temperature on chromaticity and color-rendering properties of trichromatic white-light sources based on light-emitting diodes,” J. Appl. Phys. 97(5), 054506 (2005). [CrossRef]
2. P. A. Levermore, A. B. Dyatkin, Z. M. Elshenawy, H. Pang, R. C. Kwong, R. Ma, M. S. Weaver, and J. J. Brown, “Phosphorescent OLEDs: Enabling Solid State Lighting with Lower Temperature and Longer Lifetime,” Proc. SID Symposium Digest. 42(72.2), 1060, (2011). [CrossRef]
3. 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. Top. Quantum Electron. 8(2), 310–320 (2002).
4. M. R. Krames, O. B. Shchekin, R. Mueller-Mach, G. O. Mueller, L. Zhou, G. Harbers, and M. G. Craford, “Status and future of high-power light-emitting diodes for solid-state lighting,” IEEE J. Displ. Technol. 3(2), 160–175 (2007). [CrossRef]
6. S. R. Lim, D. Kang, O. A. Ogunseitan, and J. M. Schoenung, “Potential environmental impacts from the metals in incandescent, compact fluorescent lamp (CFL), and light-emitting diode (LED) bulbs,” Environ. Sci. Technol. 47(2), 1040–1047 (2013). [CrossRef] [PubMed]
8. R. J. Xie, N. Hirosaki, N. Kimura, K. Sakuma, and M. Mitomo, “2-phosphor-converted white light-emitting diodes using oxynitride/nitride phosphors,” Appl. Phys. Lett. 90(19), 191101 (2007). [CrossRef]
9. C. C. Lin, Y. S. Zheng, H. Y. Chen, C. H. Ruan, G. W. Xiao, and R. S. Liu, “Improving Optical Properties of White LED Fabricated by a Blue LED Chip with Yellow/Red Phosphors,” J. Electrochem. Soc. 157(9), H900–H903 (2010). [CrossRef]
10. Y. H. Lin, J. P. You, Y. C. Lin, N. T. Tran, and F. G. Shi, “Development of high-performance optical silicone for the packaging of high-power LEDs,” IEEE Trans. Compon. Packag. Tech. 33(4), 761–766 (2010). [CrossRef]
11. C. C. Tsai, J. Wang, M. H. Chen, Y. C. Hsu, Y. J. Lin, C. W. Lee, S. B. Huang, H. L. Hu, and W. H. Cheng, “Investigation of Ce:YAG doping effect on thermal aging for high-power phosphor-converted white-light-emitting diode,” Trans. Device,” Mater. Res. 9(3), 367–371 (2009).
12. J. Wang, C. C. Tsai, W. C. Cheng, M. H. Chen, C. H. Chung, and W. H. Cheng, “High thermal stability of phosphor converted white light-emitting diodes employing CeYAG-doped glass,” IEEE J. Sel. Top. Quantum Electron. 17(3), 741–746 (2011). [CrossRef]
13. S. Fujita, A. Sakamoto, and S. Tanabe, “Luminescence Characteristics of YAG Glass–Ceramic Phosphor for White LED,” IEEE J. Sel. Top. Quantum Electron. 14(5), 1387–1391 (2008). [CrossRef]
14. H. Segawa, H. Yoshimizu, N. Hirosaki, and S. Inoue, “Fabrication of silica glass containingyellow oxynitride phosphor by the sol-gel process,” Sci. Technol. Adv. Mater. 12(3), 034407 (2011). [CrossRef]
15. L. Yang, M. Chen, Z. Lv, S. Wang, X. Liu, and S. Liu, “Preparation of a YAG:Ce phosphor glass by screen-printing technology and its application in LED packaging,” Opt. Lett. 38(13), 2240–2243 (2013). [CrossRef] [PubMed]
16. H. Segawa, S. Ogata, N. Hirosaki, S. Inoue, T. Shimizu, M. Tansho, S. Ohki, and K. Deguchi, “Fabrication of glasses of dispersed yellow oxynitride phosphor for white light-emitting diodes,” Opt. Mater. 33(2), 170–175 (2010). [CrossRef]
17. H. P. Fan, H. S. Chiao, H. H. Chang, and K. L. Ma, CN patent 101723586 B (2011) (in Chinese).
18. Y. K. Lee, J. S. Lee, J. Heo, W. B. Im, and W. J. Chung, “Phosphor in glasses with Pb-free silicate glass powders as robust color-converting materials for white LED applications,” Opt. Lett. 37(15), 3276–3278 (2012). [CrossRef] [PubMed]
19. C. C. Tsai, W. C. Cheng, J. K. Chang, L. Y. Chen, J. H. Chen, Y. C. Hsu, and W. H. Cheng, “Ultra-high thermal-stable glass phosphor layer for phosphor-converted white light-emitting diodes,” IEEE J. Displ. Technol. 9(6), 427–432 (2013). [CrossRef]
20. L.-Y. Chen, W.-C. Cheng, C.-C. Tsai, Y.-C. Huang, Y.-S. Lin, and W.-H. Cheng, “High-performance glass phosphor for white-light-emitting diodes via reduction of Si-Ce3+:YAG inter-diffusion,” Opt. Mater. Express 4(1), 121–128 (2014). [CrossRef]
21. L. Y. Chen, J. K. Chang, Y. R. Wu, W. C. Cheng, J. H. Chen, C. C. Tsai, and W. H. Cheng, “Optical model for novel glass based phosphor converted white light emitting diodes,” IEEE J. Displ. Technol. 9(6), 441–446 (2013). [CrossRef]
22. Y. Ohno, “Optical metrology for LEDs and solid state lighting,” Proc. SPIE 6046, 604625 (2006). [CrossRef]