Preparation of Cerium-Activated GAG Phosphor Powders
Influence of Co-doping on Crystallinity and Luminescent Properties
Chien-Chih Chiang,a,zMing-Shyong Tsai,band Min-Hsiung Honaa
Department of Materials Science and Engineering, National Cheng Kung University, Tainan 701, Taiwan b
Department of Chemical and Materials Engineering, Southern Tainan University, Yung-Kang City, Tainan 710, Taiwan
An impure-phase GdAlO3usually remains in the product of gadolinium aluminum garnet共GAG兲 powder as synthesized. In this study, an attempt to prepare a stable GAG pure-phase powder by substituting small cations共Tb3+, Y3+, or Lu3+兲 at the dodeca-hedral sites or the substituting the large cation共Ga3+兲 for the Al3+sites of the garnet structure was made. Pure garnet-phase GAG powder was formed by calcining at 1500°C for 2 h. It was found that increasing the Lu3+content in the Gd3+lattice site of the dodecahedral or increasing the Ga3+content in the Al3+site induces a blue shift in the emission wavelength. The color temperature of the pure GAG:Ce共YAG:Ce兲 phosphor powder 共⬃2900 K兲 formed was lower than that of yttrium aluminum garnet and terbium aluminum garnet共TAG:Ce兲 phosphors.
© 2007 The Electrochemical Society. 关DOI: 10.1149/1.2768900兴 All rights reserved.
Manuscript submitted March 21, 2007; revised manuscript received June 14, 2007. Available electronically August 16, 2007.
Garnet crystal is widely used at high temperatures because of its low creep rate and high thermal and chemical stabilities.1,2It is also used in solid-state laser host materials, refractory coating fillers, magnetic materials, and phosphor powders.3,4
The optical properties of Ce3+-activated garnet series phosphor used in white light emitting diodes共WLEDs兲 have received consid-erable attention in recent years.5In some illuminant applications, especially indoor illumination, the light source requires a higher color rendering index共CRI兲 and a lower color temperature to create a pleasant, relaxing atmosphere. Unfortunately, WLED using yt-trium aluminum garnet 共YAG:Ce兲 as a light-conversion layer can create a “stark” or “cold” atmosphere because of their poor color qualities. Blending other red-emitting phosphors into the YAG:Ce phosphor or changing the YAG structure by substituting cations with a different radius have been proposed as solutions to this problem.6,7 Recent researches have developed a novel garnet phosphor, cerium-activated terbium aluminum garnet共TAG:Ce兲, to produce a yellow-ish phosphor combined with a redder component.8,9Huh et al. found that theemof phosphor shifts to a longer wavelength as the doping concentration of Gd3+in the YAG:Ce structure increases.10
The unit cell of the garnet phase共兵A其3共B兲2关C兴3O12兲 has a sym-metrical cubic structure in which A, B, and C are at the dodecahe-dral, octahedral, and tetrahedral sites, respectively. In the Re2O3–Al2O3 garnet system 共Re, rare earth兲, Re ions occupy the 24共C兲 dodecahedral site 共distorted cubic lattice兲 with coordination number 共CN兲 = 8.11,12 In the Gd2O3–Al2O3 phase diagram, there are three stable compounds: gadolinium aluminum monoclinic 共GAM, Gd4Al2O9兲, gadolinium aluminum perovskite 共GAP, GdAlO3兲, and gadolinium aluminum garnet 共GAG, Gd3Al5O12兲.13In previous studies, GAG powder was prepared by quenching the glass-crystallized product from a PbO or PbF2 flux at a high tem-perature; however, the purity of the product was not described.14,15 Matyusheuko et al. indicated that it is impossible to prepare GAG powder by the solid-state method.16Cockayne et al. attempted to use the melt-growth method to prepare Gd-rich YAG, but the work was unsuccessful.17Forming a pure-phase GAG powder seems difficult. In order to form a pure garnet structure, Tb3+, Y3+, or Lu3+were adopted to substitute for Gd3+. In another process, Ga3+was adopted to substitute for Al3+. The structures and the photoluminescence properties of Tb3+, Y3+, Lu3+, and Ga3+, when substituted for GAG:Ce phosphor powder, are discussed below.
Experimental
The starting solution was Al共NO3兲3, Gd2O3, and CeCl3 at the molar ratio of Gd:Al/Ce = 2.97:5:0.03 with the total concentration
of the cation 0.04 M. A precipitation process, with 0.2 M ammo-nium hydrogen carbonate 共AHC兲 as the precipitation agent, fol-lowed the normal strike method as reported in detail in our previous studies.18,19The doping concentrations of Tb3+, Y3+, Lu3+, and Ga3+ were at 10–30 atom % 共the starting materials were Tb4O7, Y2O3, Lu2O3, and Ga2O3, respectively兲. The precipitates were filtered and dried at 100°C for 12 h and were then calcined at the desired tem-perature for 2 h. The calcined products were heated in an atmo-sphere of 95 vol % nitrogen and 5 vol % hydrogen at 1500°C for 2 h. The structure of the final products was identified using X-ray diffraction共XRD, Rigaku MultiFlex兲 with a scan speed of 4°/min and a scan step of 0.01°. Thermal behavior was determined by thermogravimetric-differential thermal analysis 共Setsys Evolution TGA-DTA兲 at a heating rate of 10°C/min under a continuous nitro-gen flow with 3 vol % hydronitro-gen. The powder morphology was ob-served using a field emission transmission electron microscope 共Hi-tachi FE2000兲. Emission and excitation spectra were detected with a Hitachi F-4500 spectrometer using a Xe lamp as the excitation source. In addition, the emission wavelength spectra of the samples were recorded with an MFS230 fluoresce spectrometer, excited at 470 nm.
Results and Discussion
Phase evolution of GAG:Ce phosphors.— Figure 1 shows the
changes in the X-ray diffraction peak intensity of the product at different calcination temperatures. Three different phases, hexagonal GAP共H-GAP兲, orthorhombic GAP 共O-GAP兲, and GAG, initially crystallized and coexisted in products after calcining at 900°C. The hexagonal perovskite structure transformed from the garnet precur-sor by wet-chemical processes20,21is easily converted into the garnet phase at higher temperatures. Yamaguchi et al.22derived a modified sol-gel process and gave the YAG synthesis scheme as
amorphous precursor→ hexagonal YAP → orthorhombic YAP
→ YAG
As presented in Fig. 1, H-GAP exists only in the temperature range below 1100°C and then transforms to GAP and GAG at higher temperatures. O-GAP is the intermediate phase at 1000–1100°C, which may be phase-transformed to the GAG phase when calcined at 1300–1400°C. Although GAG increased as the calcination temperature increased, it transformed to the O-GAP phase at temperatures above 1500–1600°C. These results indicate that the GAG phase only exists in a temperature region of
zE-mail: [email protected]
Journal of The Electrochemical Society, 154共10兲 J326-J329 共2007兲
0013-4651/2007/154共10兲/J326/4/$20.00 © The Electrochemical Society
J326
1200–1400°C. In our work, the formation of the GAG mechanism could follow hexagonal GAP 900°C → orthorhombic GAP 1500–1600°C 1000–1100°C GAG1200–1400°C Figure 2 shows the TGA/DTA results of the as-precipitated pow-der, in which most of the weight loss occurred below 880°C. The initial endothermic peak located at 150°C would be the result of dehydratation of H2O共absorbed moisture and included molecular water兲 and the decomposition of ammonium dawsonite, which is commonly observed in the AHC precipitation process.23,24Two exo-thermic peaks at about 910 and 930°C indicate the crystallization of H-GAP and GAP, respectively. The distinct exothermic peak around 1060°C may be attributed to the transformation of the O-GAP to GAG structure.
To prepare the pure garnet structure, four kinds of trivalent cat-ions, Tb3+ 共104 pm兲, Y3+ 共102 pm兲, Lu3+ 共97 pm兲, and Ga3+ 共CN:4 = 47 pm, CN:6 = 55 pm兲 were doped to substitute for Gd3+ 共106 pm兲 and Al3+共CN:4 = 39 pm, CN:6 = 53 pm兲, respectively.25 Table I is the summary of the phases presented by XRD analysis of the products calcined at 1300–1500°C. It was found that a decrease in the average substitutional ionic radius on the dodecahedral site favors a decrease in the diffraction peaks of GAP and the formation of pure GAG powder. Cockayne et al. and Kaminskii et al. reported that a single crystal of the garnet structure can be grown when the average ionic radius on the dodecahedral site is less than 0.103 nm.17,26 Kimura et al. obtained pure DyAG by co-doping larger ions, Gd3+, and smaller ions, Y3+, so that the mean radius of the dodecahedral site was controlled to less than 0.103 nm.27 Be-cause Ga3+was adopted to substitute for the Al3+site, the pure GAG phase was obtained, and it was found that an increase in Ga3+ con-tent favors the formation of pure GAG structure. The smaller ions 共Tb3+, Y3+, and Lu3+兲 substituted for the Gd3+site or the bigger ions
共Ga3+兲 substituted for the Al3+site led to a decrease in the formation temperature of the pure garnet phase. Increasing the doping concen-tration also led to a decrease in the formation temperature of the pure garnet phase.
Figure 3 shows the morphology of the as-precipitated powder and the products as heat-treated at the different temperatures. The as-precipitated powder is 10–20 nm agglomerated particles in which the primary particles are uniform and spherical. The product powder had elliptic, dense, submicrometer particles that transformed to an irregular form upon calcining at 1600°C for 2 h.
Photoluminescence properties of GAG:Ce phosphors.— The
emission wavelength of products doped with different concentra-tions of Lu3+ is shown in Fig. 4. Upon excitation at 470 nm, the emission band centered at 564.57 nm belongs to the pure GAG:Ce. Using the small cation, Lu3+, to substitute for Gd3+on the dodeca-hedral site induces the emission wavelength blue-shift. Similar re-sults were obtained by Holloway and Kestigian,28who used Lu3+to substitute for the Y3+site in YAG phosphor powder. When Gd3+or La3+was used to substitute for the Y3+site in the YAG:Ce lattice, the emission peak of the phosphor powder shifted to a longer wave-length. The peak position of the emission band depends on the Ce3+ in the garnet host environment, which is explained by considering the ionic relationship of Re–O–Ce共Re = Gd-doped Lu兲. The ionic-ity of the Re–O band decreases when Lu doping concentration in Gd increases because of the electronegativity;Lu共1.27兲 is larger than Gd共1.20兲.29In the GAG lattice near the activator center, if a rela-tively large electronegativity ion such as Lu3+is present in the GAG phosphor powder, the ionicity of Lu–O is decreased共the ionicity of Lu–O is smaller than that of Gd–O兲, which induces the increase in the ionicity of O–Ce in Lu–O–Ce. The same color shift is also observed in the Ga3+-doped GAG powder. Considering the ionic relationship of Me–O–Ce共Me = Al-doped Ga兲, the electronegativ-ity Ga共1.81兲 is larger than Al共1.61兲. The ionicity of Me–O is decreased with the increase of Ga content, which enhances the en-Figure 1. Relative intensity of the XRD peak of H-GAP共102兲, O-GAP
共121兲, and GAG 共420兲 as calcined at various temperatures.
Figure 2. DTA/TGA curves of the as-precipitated powders dried at 100°C.
Table I. Phases of the trivalent ions (Tb3+, Y3+, Lu3+, and Ga3+) substituted at the desired concentration (10, 20, and 30%) for calcined product and temperature.a Concentration Temp.共ⴰC兲 Lu10% Lu20% Lu30% Y10% Y20% Y30% Tb10% Tb20% Tb30% Ga10% Ga20% Ga30% 1500 G G G G + Pw G G G + Ps G G G G G 1400 G G G G + Pw G G G + Ps G + Pw G G G G 1300 G G G G + Ps G G G + Ps G + Ps G + Ps G + Pw G G
a“P”: GAP; “G”: GAG; subscripts w and s describe the weak and strong peaks, respectively.
J327
Journal of The Electrochemical Society, 154共10兲 J326-J329 共2007兲 J327
ergy gap of O–Ce in Me–O–Ce. Therefore, the blue shift of the emission band of Lu3+and Ga3+substituted for GAG powder occurs
as shown in Fig. 4, which means a higher energy Ce3+4f-5d absorp-tion due to the increase in O–Ce ionicity. The photoluminescence intensity of Lu3+and Ga3+doped and undoped GAG phosphor ex-cited at 470 nm is shown in Fig. 5. The phase purity also has an effect on the luminescence properties of the phosphor. Pan30 indi-cated that the decrease in the luminescence intensity of garnet phor in the Gd substituted for the dodecahedral site of YAG phos-phor is probably due to the presence of the GAP structure. Our previous report19concluded that the emission intensity of the pure YAG phase is higher than that of materials with a perovskite struc-ture 共YAP兲. The emission intensity of the undoped product of GAG:Ce is lower than that of the Lu3+- or Ga3+-doped GAG:Ce phosphor. This may be due to the presence of a perovskite structure in the GAG:Ce structure.
When comparing the luminescence property of pure GAG:Ce with that of TAG:Ce and YAG:Ce, an energy schematic diagram for Ce3+ in a garnet-series phosphor including YAG, TAG, and pure GAG can be drawn, as in Fig. 6. The 4f-5d transition of Ce3+at the ground state is excited to the conduction band and releases to the lowest 5d level. The energy then returns to 4f state and emits light. The Ce3+in the 4f state共ground state兲 splits up in a doublet state 共2F5/2,2F7/2兲 due to spin-orbit coupling. The energy difference be-tween these split states is about 2200 cm−1.31,32In the garnet-series Figure 3. SEM morphologies of共a兲 as-precipitated powders, and the
prod-ucts calcined at共b兲 900, 共c兲 1100, 共d兲 1300, 共e兲 1400, 共f兲 1500, and 共g兲 1600°C.
Figure 4. The maximum emission band of GAG/Ce powders at different
substituted concentrations of Lu3+共solid line兲 and Ga3+共dashed line兲 excited at 470 nm.
Figure 5. The photoluminescence intensity of Lu3+and Ga3+ doped and undoped GAG phosphor excited at 470 nm.
Figure 6. The energy schematic diagram of the YAG, TAG, and pure GAG
phosphor and the energy transfer process in Ce3+共*兲 Ref. 19, 共**兲 Ref. 18, and共***兲 in this report.
J328 Journal of The Electrochemical Society, 154共10兲 J326-J329 共2007兲
J328
phosphor, the emission spectra of Ce3+are quite similar. The only difference is in the wavelength of the emission peak of the host. The emission wavelength of pure GAG:Ce phosphor powder is longer than that of TAG:Ce and YAG:Ce phosphor powders because the decrease in energy for electron transferring from O2−to Ce3+ repre-sents a decrease in the ionicity of the garnet host.
The color properties of a white-light source are described by a blackbody locus共Planckian locus兲 and by the 共x,y兲 coordinates of a CIE diagram. The light sources whose共x,y兲 coordinates are situated at the blackbody locus are perceived by human eyes as white light. The CIE chromatics of YAG:Ce, TAG:Ce, and pure GAG:Ce are presented in Fig. 7. The color 共x,y兲 coordinations of YAG:Ce, TAG:Ce, and pure GAG:Ce are共0.429, 0.553兲, 共0.457, 0.529兲, and 共0.508, 0.486兲, respectively. The white light generated from WLEDs combines the blue light emitted from the LED chip and the yellow-ish phosphor. The wavelength of the blue light source is 460 nm, with the chromaticity point located at共0.144, 0.03兲. The color tem-peratures of YAG:Ce, TAG:Ce, and pure GAG:Ce are 8500 K 共0.291, 0.300兲, 4900 K 共0.348, 0.354兲, and 2900 K 共0.444, 0.406兲 as shown in Fig. 7, respectively. Therefore, for indoor illumination, pure GAG:Ce phosphor powder is more suitable as it is able to produce a warm WLED.
Conclusion
The GAG:Ce-series phosphor was successfully prepared using the coprecipitation method and calcination at certain temperatures.
The Ga3+ substituted for Al3+ and the smaller cations, including Y3+, Tb3+, and Lu3+substituted at the dodecahedral site, enhances the phase stability of GAG:Ce to form a pure garnet structure. The emission wavelength of GAG:Ce was shifted to the blue range by increasing the doping concentration of Lu3+and Ga3+. The GAG:Ce phosphor powder shows a lower color temperature when compared with YAG and TAG phosphors and thus is more suitable for warm WLED applications.
National Cheng Kung University assisted in meeting the publication costs of this article.
References
1. G. S. Corman, Ceram. Eng. Sci. Proc., 12, 1745共1991兲.
2. T. A. Parthasarathy, T. Mah, and K. Keller, Ceram. Eng. Sci. Proc., 12, 1767 共1991兲.
3. G. N. Morscher, K. C. Chen, and K. S. Mazdiyasni, Ceram. Eng. Sci. Proc., 14, 181共1994兲.
4. P. Schlotter, R. Schmide, and J. Schneider, Appl. Phys. A, 64, 417共1997兲. 5. S. Nakamura and G. Fasol, The Blue Laser Diode, GaN Based Light Emitters and
Lasers, p. 216, Springer, Berlin共1997兲.
6. R. J. Xie, N. Hirosaki, K. Sakuma, Y. Yamamoto, and M. Mitomo, Appl. Phys.
Lett., 84, 5404共2004兲.
7. K. Sakuma, K. Omichi, N. Kimura, M. Ohashi, D. Tanaka, N. Hirosaki, Y. Yama-moto, R. J. Xie, and T. Suehiro, Opt. Lett., 29, 2001共2004兲.
8. F. Kummer, F. Zwaschka, A. Ellens, A. Debray, and G. Waitl, Int. Pat. Appl. No. WO01/08452共2002兲.
9. A. Ellens and F. Zwaschka, German Pat. Appl. No. DE199 51 790 A 1共2001兲. 10. Y. D. Huh, Y. S. Cho, and Y. R. Do, Bull. Korean Chem. Soc., 23, 1435共2002兲. 11. E. Antic-Fidancev, J. Hölsä, M. Lastusaari, and A. Lupei, Phys. Rev. B, 64, 195108
共2001兲.
12. W. Y. Ching and Y. N. Xu, Phys. Rev. B, 59, 12815共1999兲. 13. P. Wu and A. D. Pelton, J. Alloys Compd., 179, 259共1992兲. 14. F. Eular and J. A. Bruce, Acta Crystallogr., 19, 971共1965兲. 15. T. Manabe and K. Egashira, Mater. Res. Bull., 6, 1167共1971兲.
16. N. N. Matyushenko, E. P. Shevyakova, E. V. Lifshits, and N. V. Lapina, Russ. J.
Inorg. Chem., 30, 942共1985兲.
17. B. Cockayne, D. B. Gasson, D. Findlay, D. W. Goodwin, and R. A. Clay, J. Phys.
Chem. Solids, 29, 905共1968兲.
18. C. C. Chiang, M. S. Tsai, and M. H. Hon, J. Alloys Compd., 431, 298共2007兲. 19. C. C. Chiang, M. S. Tsai, C. S. Hsiao, and M. H. Hon, J. Alloys Compd., 416, 265
共2006兲.
20. P. S. Devi, Y. Lee, J. Margolis, J. B. Parise, S. Sampath, H. Herman, and J. C. Hanson, J. Mater. Res., 17, 2846共2000兲.
21. S.-M. Sim, K. A. Keller, and T.-I. Mah, J. Mater. Sci., 35, 713共2000兲. 22. O. Yamaguchi, K. Takeoka, K. Hirota, H. Takano, and A. Hayashida, J. Mater. Sci.,
27, 1261共1992兲.
23. K. Hayashi, S. Toyoda, K. Nakashima, and K. Morinaga, J. Ceram. Soc. Jpn., 98, 444共1990兲.
24. J. G. Li, T. Ikegami, J. H. Lee, and T. Mori, J. Mater. Res., 15, 2375共2000兲. 25. R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B: Struct. Crystallogr.
Cryst. Chem., 25, 925共1969兲.
26. A. A. Kaminskii, A. G. Petrosyan, K. L. Ovanesyan, G. O. Shironyan, V. A. Fedorov, and S. E. Sarkisov, Phys. Status Solidi A, 82, K185共1984兲.
27. H. Kimura, T. Numazawa, M. Sato, H. Maeda, and M. Sakamoto, J. Cryst. Growth,
97, 607共1989兲.
28. W. W. Holloway and M. Kestigian, J. Opt. Soc. Am., 59, 60共1969兲. 29. A. L. Allred, J. Inorg. Nucl. Chem., 17, 215共1961兲.
30. Y. Pan, M. Wu, and Q. Su, Mater. Sci. Eng., B, 106, 251共2004兲. 31. Q. Li, L. Gao, and D. Yan, Mater. Chem. Phys., 64, 41共2000兲.
32. J. M. Robertson, M. W. van Tol, W. H. Smits, and J. P. H. Heynen, Philips J. Res.,
36, 15共1981兲.
Figure 7.共Color online兲 The color coordination 共x,y兲 and color temperature
for YAG共䉭兲, TAG 共쎲兲, and pure GAG 共䊊兲 phosphor powders in CIE diagram.
J329
Journal of The Electrochemical Society, 154共10兲 J326-J329 共2007兲 J329
