Japanese Journal of Applied Physics 48 (2009) 032001

REGULAR PAPER

Photoluminescence Characteristics of Macroporous Eu-Doped Yttrium Oxide Particles Prepared by Spray Pyrolysis W. Widiyastuti, Takaaki Minami1 , Wei-Ning Wang, Ferry Iskandar, and Kikuo Okuyama Department of Chemical Engineering, Graduate School of Engineering, Hiroshima University, Higashihiroshima, Hiroshima 739-8527, Japan 1 Scientific Instruments and Systems Division, Horiba Ltd., Kyoto 601-8510, Japan Received September 30, 2008; accepted December 17, 2008; published online March 23, 2009 The photoluminescence (PL) characteristics of ordered macroporous europium-doped yttrium oxide (Y2 O3 :Eu3þ ) particles were investigated. The submicrometer particles were prepared by spray pyrolysis using a mixture of a yttrium and europium nitrate solution and colloidal polystyrene latex (PSL) particles as the precursor. The porous particles exhibited higher PL intensity, quantum efficiency, and red-emission properties than the non-porous particles due to their porous structures. Detailed PL analysis revealed that the porous particles have unique photophysical properties, such as a 4-nm blue-shift in the charge transfer band (CTB) wavelength, and a lower symmetry ratio of 2.7, which were not found in non-porous particles with approximately the same particle and crystallite sizes. # 2009 The Japan Society of Applied Physics DOI: 10.1143/JJAP.48.032001

1.

Introduction

Rare earth-doped phosphor materials have been used in many applications, such as field emission displays, cathode ray tubes, and plasma display panels. Among these materials, europium-doped yttrium oxide (Y2 O3 :Eu3þ ) is one of the most promising red phosphors due to its stability under vacuum, high thermal conductivity, and high melting point.1,2) Mostly non-porous Y2 O3 :Eu3þ particles, i.e., dense/filled particles, can be prepared using a variety of methods, and the corresponding photoluminescence (PL) characteristics of the particles with regard to different factors, such as morphology, particle size, crystallinity, and composition, have been investigated.1–5) In contrast, porous materials recently have attracted increasing attention for applications in electrochemistry,6) nanomaterials,7,8) photonic crystals,9) and drug delivery for tracking the efficiency of drug release owing to their high surface area, low density, and hierarchical structures.10) Similar to nanomaterials, ordered porous particles maintain material properties on the submicrometer scale. However, ordered porous particles have advantages over nanomaterials: superior surface and thermal stability, easy handling and shape preservation.11) Recently, our group reported the preparation of various porous materials, such as silica (SiO2 ), titania (TiO2 ), alumina (Al2 O3 ), zirconia (ZrO2 ), and yttria (Y2 O3 ) particles, using a spray pyrolysis method. This method is promising for the following reasons: it is continuous, rapid, low cost; and the ability to control stoichiometry, size and product composition is good.12,13) In the spray method, precursors of inorganic salt solutions mixed with colloidal polystyrene latex (PSL), as templates, are often used to prepare porous particles. The porosity and pore size can be easily controlled by changing the PSL content and size, respectively.14,15) The synthesis of other porous particles or films have been extensively investigated by many groups using various methods, such as dip-coating and sol gel methods.16,17) Recent progress in the various methods available for production of macroporous particles was reviewed by Studart et al.18) 

E-mail address: [email protected]

However, to the best our knowledge, few studies have investigated the preparation and characterization of macroporous phosphors, in particular, the rare earth-doped oxide phosphors. In this study, the PL characteristics of macroporous phosphor Y2 O3 :Eu3þ particles synthesized using a spray pyrolysis method were investigated. The effects of porosity and particle size on the PL characteristics, including PL intensity, charge transfer peak shift, symmetry ratio, quantum efficiency (QE), and XY CIE chromaticity values were examined in detail. 2.

Experimental Procedure

The spray pyrolysis experimental setup used in the present study was described in detail in our previous work.1,19) Yttrium nitrate [Y(NO3 )3 6H2 O)] and europium nitrate [Eu(NO3 )3 6H2 O] (Kanto Chemical) were used as precursors and were dissolved in ultrapure water to form homogeneous aqueous solutions. Europium atoms were added to a doping level of 6 mol % relative to yttrium atoms. To prepare porous yttrium oxide particles, yttrium and europium nitrate aqueous solutions were mixed with 300-nm PSL colloids (Japan Synthetic Rubber). The content of the PSL (vol %) added to the precursor, which represents ‘‘porosity’’ in this paper, was calculated based on the volume ratio of PSL/Y2 O3 in the final product particles. The precursor solutions were atomized into droplets using an ultrasonic nebulizer (Omron Healthcare NE-U17). The mean droplet diameters of the pure yttrium nitrate solution and the mixture of yttrium nitrate and PSL were 5.12 and 5.18 mm, respectively, based on laser diffraction measurements (Malvern Instruments Spraytec). This result indicates that the presence of PSL colloids in the precursor does not significantly affect the droplet size distribution. The droplets were then carried into a tubular alumina reactor by nitrogen gas at a flow rate of 2 L/min. The tubular reactor was 1 m in length with an inner diameter of 13 mm. The temperatures of five controllable furnace zones were set to 700, 1000, 1400, 1400, and 1400  C. Prior to entering the furnace, the droplets passed through a 20-cm long zone heated to 160  C to evaporate water. The resulting particles were collected in an electrostatic precipitator. The morphology of the synthesized particles was observed using a field emission scanning electronic microscope (FESEM; Hitachi S-5000) operated at 20 kV. The crystallinity

032001-1

.

.

# 2009 The Japan Society of Applied Physics

Jpn. J. Appl. Phys. 48 (2009) 032001

W. Widiyastuti et al.

was analyzed using X-ray diffraction (XRD; Rigaku RINT 2200V) with Cu K radiation in the range of 10 – 80 at 40 kV and 30 mA. PL properties were examined using a spectrofluorophotometer (Shimadzu RF-5300PC) with an excitation wavelength of 254 nm, a bandwidth resolution less than 3 nm, and a signal/noise (S/N) ratio greater than 150. Excitation spectra were recorded under emission at 611 nm using the same equipment. Quantum efficiency (QE) and CIE chromaticity measurements (Hamamatsu C992002) were performed to analyze PL characteristics in detail. All measurements were carried out at room temperature. 3.

Sample

C (M)

PSL (vol %)

dp (nm)

dc (nm)

QE (%)

CIE coordinates X

Y

Y-1

0.5

0

796

29.3

60

0.632

0.347

Y-2

0.3

40

774

31.6

74

0.642

0.343

Y-3 Y-4

0.2 0.15

60 70

768 769

31.6 31.6

76 77

0.642 0.642

0.342 0.342

Y-5

0.1

80

757

30.4

79

0.641

0.341

Y-6

0.0375

80

650

29.3

65

0.642

0.340

Y-7

0.15

80

886

34.2

80

0.641

0.340

Results and Discussion

To investigate the effect of porosity, particle size was calculated and kept constant based on the typical one droplet to one particle principle (ODOP),19) by changing the concentrations of yttrium and europium nitrate aqueous solutions and the corresponding PSL contents from 0.5, 0.3, 0.2, 0.15 to 0.1 M, and 0, 40, 60, 70 to 80 vol %, respectively. On the other hand, to examine the effect of particle size, the precursor concentration was altered from 0.0375, 0.1 to 0.15 M, while the PSL content was maintained at 80 vol %. The experimental conditions are summarized in detail in Table I. 3.1

Table I. Effects of porosity and particle size on crystallite size, QE and CIE coordinates.

Characterization of morphology, particle size, and crystallinity Figure 1 shows FE-SEM images of the particles prepared from precursors with different PSL contents with approximately the same particle size, i.e., 770 nm. Non-porous particles were produced from the precursor without the use of PSL colloids for comparison. As shown in Fig. 1(a), the particles were spherical and dense, as confirmed by computing the particle size based on the one droplet to one particle (ODOP) principle.19) In Fig. 1(b), hollow rather than porous particles were generated due to the low PSL content (40 vol %), which tended to migrate to the center of droplets before decomposition at high temperatures. Increasing the PSL content increased the pore number, as shown in Figs. 1(c) and 1(d). With addition of 80 vol % PSL to the precursor, pores were seen both inside the particles and also on the particle surface. This result indicates that the PSL particles were spontaneously well-dispersed during droplet evaporation. The effect of porous particle size was also investigated by maintaining constant porosity with addition of 80 vol % PSL colloids to the precursors. Particle size was controlled by varying the concentration of the nitrate solution, i.e., 0.0375, 0.1, and 0.15 M, resulting in average particle sizes of 650, 757, and 886 nm, respectively, as measured from the FESEM images shown in Fig. 2. XRD patterns of the generated particles are presented in Fig. 3. All particles were characterized as pure bodycentered-cubic (bcc) Y2 O3 phase corresponding to JCPDS No. 41-1105 with no impurities nor phase transition. In addition, clear peaks were observed for all particles, indicating that the selected furnace temperature profile was effective for the synthesis of both dense and porous particles with relatively high crystallinity. The crystallite sizes were calculated using the Scherrer equation based on full width at half maximum (FWHM) values at 2 ¼ 29:1 . The crystal-

lite size was 29.3 nm for the non-porous particles. For most porous particles that were similarly sized as non-porous particles, crystallite size was relatively constant at approximately 31.6 nm. For particles with higher porosity synthesized from the precursor with 80 vol % PSL, crystallite size declined slightly to approximately 30.4 nm. The slightly higher crystallinity in porous particles might be due to the relatively lower inorganic salt content in the precursor under the same heat treatment. However, in the case of particles with high porosity, more energy is required for evaporation/ decomposition of the PSL particles (e.g., 80 vol %), resulting in a decrease in crystallite size. The crystallite size of different sized particles was also shown by the XRD patterns in Fig. 3. It is evident that increasing particle size from 650 to 886 nm by increasing the precursor concentration from 0.0375 to 0.15 M with the same heat treatment leads to an increase in crystallite sizes from 29.3 to 34.2 nm. This phenomenon occurs because crystallite growth is affected by the number concentration of Y2 O3 primary particles (nuclei) and the solvent concentration that must be evaporated. At equivalent droplet energy, the nucleation occurs earlier for more concentrated precursors, consequently the time available for crystal growth is longer, resulting in higher crystallinity. Another reason that particles prepared from less concentrated precursors is that less energy is available for crystal growth, as more energy is needed for solvent evaporation. 3.2 Photoluminescence spectra Figure 4 shows PL emission spectra of Y2 O3 :Eu3þ particles with different porosities (i.e., different PSL contents) (a) and particle sizes (b), and magnified PL spectra from 570 to 600 nm for both porous (80 vol %) and non-porous particles. A strong emission peak at 611 nm in the PL emission spectra is due to the 5 D0 to 7 F2 allowed electrical dipole transition of Eu3þ . The intensity of the main emission peak increased with increasing porosity, i.e., PSL content. The relative PL intensity of the strongest emission peak at 611 nm and crystallite size as a function of porosity are shown in Fig. 5(a). The PL intensity of non-porous particles was used as the basis (100%) for comparison. As shown in the figure, the relative PL intensity increased significantly from nonporous particles to porous particles with addition of 40 vol % PSL. The relative PL intensity for porous particles continued to increase to 133% that of non-porous particles with addition of 70 vol % PSL. In the case of porous particles with addition of 80 vol % PSL, approximately 135% relative PL

032001-2

# 2009 The Japan Society of Applied Physics

Jpn. J. Appl. Phys. 48 (2009) 032001

(a)

W. Widiyastuti et al.

(c)

(b)

Fig. 1.

300nm

300nm

300nm

300nm

1.20μm

1.20μm

1.20μm

1.20μm

FE-SEM images of particles prepared from precursors with PSL contents of (a) 0, (b) 40, (c) 60, and (d) 80 vol %.

(a)

(c)

(b)

1.20μm

1.20μm

1.20μm

Intensity [arb. unit]

(211)

(222)

FE-SEM images of particles prepared from precursors with different concentrations of 80 vol % PSL (a) 0.0376, (b) 0.1, and (c) 0.15 M.

JCPDS 41-1105 Y2O3, Cubic

(400) (411) (332) (431) (440) (611) (622)

Fig. 2.

(d)

PSL: 0 vol% dp=796nm

dc=29.3nm

PSL: 40 vol% dp=774nm

dc=31.6nm

PSL: 60 vol% dp=768nm

dc=31.6nm

PSL: 70 vol% dp =769nm

dc=31.6nm

PSL: 80 vol% dp=757nm

dc=30.4nm

PSL: 80 vol% dp=650nm

dc=29.3nm

PSL: 80 vol% dp =886nm

dc=34.2nm

10

20

30

40

50

60

70

80

2θ [deg] Fig. 3. XRD patterns of particles with different porosities and particle sizes.

intensity was achieved. The crystallite size also tended to increase with increasing the porosity and then it maintains a relatively constant value with 40 – 70 vol % PSL in the

precursor. The crystallite size decreased for porous particles that were synthesized using >70 vol % PSL in the precursor, as explained above. Based on the results of the present study, it is obvious that crystallinity is not the main reason for the greater PL intensity of the porous particles. Therefore, the primary explanation for this phenomenon is their porous structure. The Y2 O3 base crystal has two symmetry types, C2 and C3i . The PL emission peaks at 587.7 nm [5 D0 (C2 ) ! 7 F1a (C2 )] and 582.4 nm [5 D0 (C3i ) ! 7 F1a (C3i )] were transformed from the C2 and C3i states, respectively as shown in Fig. 4(c). The ratio of the PL emission peak intensity at 587.7 nm to that at 582.4 nm is defined as the symmetry ratio of the C2 to C3i sublattices, which is the basic parameter by which PL emission spectra are evaluated. As stated in our previous paper, a symmetry ratio of approximately 3 indicates that Eu3þ does not preferentially occupy the C2 or C3i sites.5) Porous particles exhibit lower symmetry than non-porous particles. For example, the symmetry ratios were calculated to be 2.7 and 3.1 for porous particles with 80 vol % PSL addition and non-porous particles of the same size, respectively. Compared with non-porous particles, porous particles have more atoms located on the particle surface due to their high surface area, and surface defects in the crystallites may occur more frequently. These defects may increase the degree of disorder and lower the local symmetry of Eu3þ ions located near the surface. The reduced local symmetry,

032001-3

# 2009 The Japan Society of Applied Physics

560

580

600

620

640

660

(b)

560

Wavelength [nm]

580

600

620

640

660

D0(C2 ) 5

7 D0 F0(C2 ) 7 D0(C3i ) F 1a(C3i )

575

5

dp =650 nm

5

d p=757 nm

porous non-porous

7

(c) d p=886 nm

Intensity [arb. unit]

Intensity [arb. unit]

PSL addition: 80 vol% 70 vol% 60 vol% 40 vol% 0 vol%

Intensity [arb. unit]

(a)

W. Widiyastuti et al.

F 1a(C2 )

Jpn. J. Appl. Phys. 48 (2009) 032001

580

Wavelength [nm]

585

590

595

600

605

Wavelength [nm]

Fig. 4. PL emission spectra of Y2 O3 :Eu3þ particles as a function of porosity plotted against volume percentages of added PSL (a) andparticle size with a constant PSL content of 80 vol % (b); and magnified PL emission spectra from 570 to 600 nm for both porous (80 vol %) and nonporous particles (c).

40 38

120

36

110

34

100

32

90

30

80

0

20

40

60

80

Intensity [arb. unit]

130

28

220

240

38

34 32 30 80 28

700

800

900

300

dp=886 nm

Intensity [arb. unit]

120

Crystallite size [nm]

Relative PL intensity [%]

36

60 600

280

(b)

(b)

100

260

Wavelength [nm]

PSL addition [vol%] 140

PSL addition: 80 vol% 70 vol% 60 vol% 40 vol% 0 vol%

(a)

(a)

Crystallite size [nm]

Relative PL intensity [%]

140

220

26

dp=757 nm dp=650 nm

240

260

280

300

Wavelength [nm]

Particle size [nm] Fig. 5. Relative PL emission peak intensity and crystallite size of the generated particles as a function of (a) porosity and (b) particle size.

in other words, the increased asymmetry, will lead to differences in the intensity ratio of Y2 O3 :Eu3þ phosphor particles, and hence increase the PL emission intensity.1,5) Figure 4(b) shows very similar PL emission spectra for Y2 O3 :Eu3þ porous particles with different particle diameters. The PL emission intensity increased significantly when the particle size was increased from 650 to 757 nm, but remained constant for particles larger than 800 nm [Fig. 5(b)]. This phenomenon was also observed and explained in our previous investigation of non-porous Y2 O3 :Eu3þ particles.1) To further investigate the PL characteristics of the porous particles, the excitation spectra were also recorded for particles with different porosities and particle sizes

Fig. 6. PL excitation spectra of Y2 O3 :Eu3þ particles as a function of porosity (a) and particle size (b).

[Figs. 6(a) and 6(b), respectively]. The broad excitation spectrum peak located at approximately 250 nm in Fig. 6(a) was assigned to the excited charge-transfer band (CTB) of Eu3þ . The CTB peak wavelength of the non-porous particles was located at 250 nm, which is the same value as that of the bulk. CTB peak wavelengths of the porous particles with addition of 40, 60, 70, and 80 vol % PSL were blue-shifted from 248, 247, 246 to 246 nm, respectively. In our previous report,5) Y2 O3 :Eu3þ particles has blue shifted CTB peak proportional to the diameter of the particles less than 600 nm. Results of our previous study revealed that the CTB peak wavelengths of Y2 O3 :Eu3þ 590 and 390 nm non-porous particles were 247 and 244 nm, respectively.5) As shown in Fig. 6(a), the porous particles with an average size of

032001-4

# 2009 The Japan Society of Applied Physics

Jpn. J. Appl. Phys. 48 (2009) 032001

W. Widiyastuti et al.

3.3 Color rendering properties The effects of porosity and particle size were also examined using QE measurements. The QE values of all samples are listed in Table I. The QE values increased with increasing porosity. For example, the QE value of non-porous particles increased significantly from 60 to 74% for the particles synthesized using addition of 40 vol % PSL to the precursor. The QE value reached 79% for particles synthesized using addition of 80 vol % PSL. In the case of particles with different sizes, but constant porosity, the QE increased with increasing particle size. Thus, the QE measurements confirmed the PL intensity measurements. The CIE1931 XY chromaticity coordinate graph was plotted to characterize the color rendering properties of the samples, as shown in Fig. 7. The CIE chromaticity coordinates of saturated red phosphors based on the standard of National Television Standard Committee (NTSC) are at x ¼ 0:67 and y ¼ 0:33.21) As shown in Table I, the chromaticity coordinates of all samples were very close to the standard x value, indicating that high quality red-emitting phosphor particles, efficient for most applications, were generated. Moreover, the quality of the red, ordered macroporous particles was slightly superior to that of the non-porous particles, even if the mass density of the macroporous particles was reduced due to high porosity. 4.

Conclusions

Ordered, macroporous, red-phosphor particles with high quantum efficiency and sharp red-emission indicated by XY CIE chromaticity were synthesized by spray pyrolysis. The porous particles had unique photophysical properties, such as a 4-nm blue-shift of CTB wavelength, and a lower symmetry ratio of 2.7, that were not shown by non-porous particles with approximately the same particle and crystallite sizes. The structure of these porous particles is considered the main reason for this phenomenon. These results suggest that the porous phosphor particles are promising for a variety of applications in advanced industries due to their high PL characteristics, low density, and hierarchical structures.

0.8

0.6

Y [-]

770 nm (with addition of 60 and 70 vol % PSL) show approximately the same CTB wavelength as the 590-nm non-porous particles. Van Pieterson et al. reported that the peak wavelength shift of the CTB by different host crystals, such as Y2 O3 , Y2 O2 S, and YPO3 , is determined by the nature of the ligand, the metal ion, the coordination, and the size of cation site.20) In Y2 O3 :Eu3þ , the CTB is closely related to the covalency between O2 and Eu3þ and the coordination environment of Eu3þ . The Eu–O distance is the primary determinant of the CTB peak position. Crystallite size has a minor influence, and particle size has a major influence, on the CTB peak shift.5) For porous particles, the blue-shift was due to the bond length of Eu–O and its covalency. For porous particles, the porous structure may be the primary determinant of the coordination environment of Eu3þ and, hence, the Eu–O bond distance. Therefore, porous particles show a CTB blue-shift. In Fig. 6(b), the PL excitation peak was maintained at approximately 246 nm, indicating that CTB did not change with particle size in this region.

Y-1

0.4

Y-2 to Y-7

0.2

0.0 0.0

0.2

0.4

0.6

X [-] Fig. 7. (Color online) CIE chromaticity diagram of Y2 O3 :Eu3þ particles with different porosities and particles sizes (samples of Y-1 to Y7), as indicated in Table I.

Acknowledgements The authors wish to thank Shunsuke Kinouchi for assistance in the experimental work. We acknowledge the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan for providing a doctoral scholarship (W.W.).

1) W.-N. Wang, W. Widiyastuti, T. Ogi, I. W. Lenggoro, and K. Okuyama: Chem. Mater. 19 (2007) 1723. 2) S. L. Jones, D. Kumar, R. K. Singh, and P. H. Holloway: Appl. Phys. Lett. 71 (1997) 404. 3) L. E. Shea, J. Mckittrick, and M. L. F. Phillips: J. Electrochem. Soc. 145 (1998) 3165. 4) N. Joffin, J. Dexpert-Ghys, M. Verelst, G. Baret, and A. Garcia: J. Lumin. 113 (2005) 249. 5) T. Minami, W.-N. Wang, F. Iskandar, and K. Okuyama: Jpn. J. Appl. Phys. 47 (2008) 7220. 6) F. Wang, R. Liu, A. Pan, S. Xie, and B. Zou: Mater. Lett. 61 (2007) 4459. 7) H. Q. Wang, G. Z. Wang, L. C. Jia, C. J. Tang, and G. H. Lie: J. Phys. D 40 (2007) 6549. 8) G. Liu and G. Hong: J. Nanosci. Nanotechnol. 6 (2006) 3442. 9) H. Yan, Y. Yang, Z. Fu, B. Yang, Z. Wang, L. Xia, S. Yu, S. Fu, and F. Li: Chem. Lett. 34 (2005) 976. 10) P. Yang, S. Huang, D. Kong, J. Lin, and H. Fu: Inorg. Chem. 46 (2007) 3203. 11) K. Okuyama, M. Abdullah, I. W. Lenggoro, and F. Iskandar: Adv. Powder Technol. 17 (2006) 587. 12) M. Abdullah, F. Iskandar, S. Shibamoto, T. Ogi, and K. Okuyama: Acta Mater. 52 (2004) 5151. 13) W.-N. Wang, F. Iskandar, K. Okuyama, and Y. Shinomiya: Adv. Mater. 20 (2008) 3422. 14) F. Iskandar, Mikrajuddin, and K. Okuyama: Nano Lett. 1 (2001) 231. 15) F. Iskandar, Mikrajuddin, and K. Okuyama: Nano Lett. 2 (2002) 389. 16) F. Iskandar, M. Abdullah, H. Yoden, and K. Okuyama: J. Sol–Gel Sci. Technol. 29 (2004) 41. 17) S. Mann, S. L. Burkett, S. A. Davis, C. E. Fowler, N. H. Mendelson, S. D. Sims, D. Walsh, and N. T. Whilton: Chem. Mater. 9 (1997) 2300. 18) A. R. Studart, U. T. Gonzenbach, E. Tervoort, and L. J. Gauckler: J. Am. Ceram. Soc. 89 (2006) 1771. 19) W.-N. Wang, W. Widiyastuti, I. W. Lenggoro, T. O. Kim, and K. Okuyama: J. Electrochem. Soc. 154 (2007) J121. 20) L. Van Piterson, M. Heeroma, E. de Heer, and A. Meijerink: J. Lumin. 91 (2000) 177. 21) Y.-H. Niu, Y.-L. Tung, Y. Chi, C.-F. Shu, J. H. Kim, B. Chen, J. Luo, A. J. Carty, and A. K.-Y. Jen: Chem. Mater. 17 (2005) 3532.

032001-5

# 2009 The Japan Society of Applied Physics