1. Introduction

Nowadays, white light emitting diodes (WLEDs), a new type of luminescent source, have received considerable attention because of their advantage in energy conservation and broad application potential over conventional illumination devices. The current leading commercial WLED normally combines an InGaN blue-emitting chip with Ce:YAG yellow-emitting powder phosphor, and the powder phosphor is packed on the chip surface by using epoxy resin [1–7]. However, the deterioration of the organic resin matrix, because of its poor heat-resistance, becomes a serious problem under the increasing temperature and output power of LED chips. This raises several issues in WLED practical application, such as degradation of luminous efficacy and long-term reliability, shifts in emission color, and reduction of lifetime.

In order to solve the problem, researchers have investigated novel durable phosphors without resins. All-inorganic phosphors are on track to replace the “phosphor powder”-“organic matrix” combination, such as transparent ceramic phosphor [8, 9] and glass ceramic phosphor as practical alternatives to solve the deterioration of the organic resin matrix. Sakata et al. invented the ceramic composite which has an irregularly twisted lamellar structure with Al₂O₃ and Ce:YAG [10]. However, it’s reported that the luminous efficacy is still not excellent [9]. In a thin transparent Ce:YAG ceramic-based WLED, a maximum luminous efficacy reaching 93 lm/W at a low correlated color temperature (CCT) of 4600 K was realized [11]. However, the Ce:YAG nanopowders need to be synthesized by co-precipitation, so the increasing cost of the complex fabrication process is an unavoidable challenge for industrial production. Notably, transparent Ce:YAG phosphor-in-glass also shows a good performance [12]. Zhang et al. reported that transparent Ce:YAG phosphor-in-glass yielded a luminous efficacy of 124 lm/W, a correlated color temperature of 6674K and a color rendering index (CRI) of 70, at an operating current of 350mA. However, the thermal conductivity of glass is far inferior to ceramic. This may limit its application in some cases, for instance, when laser is adopted as the excitation light source.

In the present study, we have made two-phase Al₂O₃-Ce:YAG ceramic phosphor with Ce³⁺-correlated yellow emission. The ceramic phosphor uniformly mixes hexagonal α- Al₂O₃ with cubic YAG. We also investigate its resulting luminescent properties. The luminous efficacy of the WLED based on the composite phase ceramic phosphor reaches 95 lm/W, superior to that of the single-phase transparent Ce:YAG ceramic.

2. Experimental methods

The ceramic samples with chemical formula a(Ce_0.001Y_0.999)₃Al₅O₁₂:bAl₂O₃ were prepared by solid-state reaction at high temperature and high purity powders Y₂O₃ (99.999%), Al₂O₃(99.999%) and CeO₂ (99.99%) were used as raw materials. 0.1 wt% MgO and 0.4wt% TEOS were used as the additives, 1wt% PEG as the dispersant [13, 14]. In the weighting process, the mole ratio of Al₂O₃/Y₂O₃ was controlled in 2~8, and CeO₂ is the dopant at a mole ratio of Ce/(Y + Ce) = 0.1%. The final ceramic component proportion is Al₂O₃/YAG = 0~10.35 in mole, 0~1.78 in weight, and 0~2.08 in volume, calculated according to their molecular weight and density.

Homogeneous slurry of the mixed raw materials was obtained by wet ball milling with ethanol for 12h, and then dried in oven at 70°C. The obtained powders were uniaxially pressed into disks at 10MPa and then isostatically cold pressed at 210 MPa. The compacted disks were sintered at 1700°C for 10 h under a vacuum of 10⁻³ Pa. Al₂O₃-Ce:YAGceramics withφ16 mm in diameter was obtained. We cut and polished the ceramics to the appropriate thickness for producing white light on the blue LED chip.

The crystalline phases of the samples were identified by XRD measurement with a Cu Kα radiation source (Ultima IV Diffractometer, Rigaku, Japan) in the angular range of 2θ = 10–90°at a scan step width of 0.02°. Elemental analysis of the Al₂O₃-Ce:YAGceramic was carried out by energy dispersive X-ray spectroscopy (EDS) under the SEM. Fracture surface microstructure was observed by the SEM (JSM-6510, JEOL, Japan). Luminous efficacy was evaluated with an integrating sphere (Everfine PMS-50 system) in Shanghai Semiconductor Lighting Engineering Research Center.

3. Results and discussion

The obtained samples of the ceramic phosphor are shown in Fig. 1. In the system which is Al₂O₃ and Y₂O₃ as raw materials, the final phases depend on the sintering temperature [15]. In our case (Fig. 2), the XRD reveals the characteristic peaks of α-Al₂O₃ and cubic-YAG, and no peaks from other crystalline phases. The samples with different molar ratios of Al₂O₃/Y₂O₃ have different relative peak intensities fromα-Al₂O₃ and cubic-YAG, and with increasing of the molar ratio of Al₂O₃/Y₂O₃, the intensity ofα-Al₂O₃ increases correspondingly. Furthermore, the SEM-EDS mapping of the ceramic phosphor also confirms only Al₂O₃ and cubic-YAG existed, as demonstrated in Fig. 3. The region marked by “Spectrum 12” is YAG phase and the area of “Spectrum 14” is Al₂O₃ phase according to EDS analyses (Table 1), which is in accordance with the XRD results mentioned above. Through the observation of the fracture surface morphology, someα- Al₂O₃ particles are embedded in the YAG matrix and the others have been extracted. Therefore, we find some convex shapes and concave impressions. Supplementary note, in the XRD and EDS there is no information to reflect the state of Ce ion because of its low concentration.

 

Fig. 1 Photographs of the Ce-doped Al₂O₃-YAG ceramics.

Download Full Size | PDF

 

Fig. 2 XRD patterns of the Ce-doped Al₂O₃-YAG ceramic.

Download Full Size | PDF

 

Fig. 3 SEM-EDS mapping of the Al₂O₃-Ce:YAG ceramic.

Download Full Size | PDF

Table 1. The SEM-EDS composition of the marked area in Fig. 3

View Table | View all tables in this article

SEM observations on the fracture surface of the samples were carried out to investigate variation of the microstructure with the mole ratio of Al₂O₃ /YAG, as show in Fig. 4. Obviously, the YAG crystal is mainly with transgranular fracture mode, the Al₂O₃partical with intergranular fracture mode, and meanwhile most ofAl₂O₃particalshave been extracted from the YAG matrix forming concave impressions. The Al₂O₃ particles sized 2-3μm are homogeneously dispersed in the YAG matrix. And with the decreasing of the mole ratio of Al₂O₃ /YAG, the amount of Al₂O₃ particles is decreasing and the transparency is changed from translucent to transparent, as seen in the inset of Fig. 4. The thicknesses of the samples in Figs. 4(a)-4(c) are 0.44mm, 0.44mm and 0.64mm, respectively. Notably, though the sample in Fig. 4(c) is much thicker than others, it exhibits the best transparency. The Al₂O₃ particles change the linear propagation of light in ceramics, therefore the in-line transmittance deceases with increasing Al₂O₃ content. We also use the thermal etching to make the YAG grain boundary exposed, and the SEM observation reflects an interesting information of YAG grain. As shown in Fig. 4(d), the grain size of YAG is 5-15μm. Comparing the result with the report by Liu [15] under the same sintering temperature, we find that the existence of Al₂O₃ don’t influence the grain size of YAG.

 

Fig. 4 Photographs of the Al₂O₃-Ce:YAG transparent ceramics prepared under various experimental components: (a). the mole ratio of Al₂O₃ /YAG = 1.46 and Ce:Y = 0.001, (b). the mole ratio of Al₂O₃ /YAG = 0.65 and Ce:Y = 0.001, (c). the mole ratio of Al₂O₃ /YAG = 0.0058 and Ce:Y = 0.0012; (d). SEM image of the thermally etched surface of the sample with the mole ratio of Al₂O₃ /YAG = 0.65 and Ce:Y = 0.001.

Download Full Size | PDF

Figure 5 exhibits WLED emission spectra with the ceramic phosphoroftheAl₂O₃/YAG mole ratio 0.65 when put on the blue chip. The CRI of the sample is 63.4 and the CCT is 5585K.The blue LED chip emits blue light with the dominant wavelength of 452.6 nm under a drive current of 150 mA, and subsequently the blue light excites the Al₂O₃-Ce:YAG ceramic phosphor to produce emission light within 500~700nm. The absorption at ~450 nm comes from the $\text{4f} \text{5d}$ transition of Ce³⁺, and the photoluminescence spectrum of the Al₂O₃-Ce:YAG ceramic exhibits a typical ${\text{Ce}}^{\text{3}+}:\text{5d} \text{4f}$ broadband emission centered at 540nm under the 450 nm excitation in the YAG host. The emission light and the blue light which are not absorbed by Ce³⁺ form white light together, as shown in Fig. 5.

 

Fig. 5 Electroluminescence spectra of WLED with the Al₂O₃-Ce:YAG ceramic phosphor.

Download Full Size | PDF

The measured luminous efficiency of WLED with the Al₂O₃-Ce:YAG ceramic phosphors is listed in Table 2 and all the samples were measured at room temperature (25°C). The Al₂O₃-Ce:YAG ceramics with different mole ratios of Al₂O₃/YAG were cut into appropriate thicknesses to emit white light when packed on the blue LED chip. When the content of YAG in weight is more than 90%, on increasing the Ce:YAG content, the thicknesses of the samples increase. Meanwhile, the transparency of the samples increases gradually, shown as Fig. 1. However, when the content of Ce:YAG in weight is less than 90%, on increasing the Ce:YAG content, the thicknesses of the samples decrease, as shown in Table 2. The Ce:YAG matrix in different ceramic phosphors has the same Ce³⁺ concentration at a mole ratio of Ce/Y = 0.1%. Thus the Ce³⁺ doping amount increases with the increasing of Ce:YAG content. Therefore, the samples turn thinner gradually to achieve the white light when packed on the blue chip [16]. A lot of Al₂O₃ particles in the Ce:YAG matrix lead to the ceramic opaque, but the Al₂O₃ particles and the grain boundary can change the direction of light propagation in ceramic, which may increase the absorption probability of the blue light (Fig. 6a). So the proper amount of Al₂O₃ particles can reduce the thickness of samples to get white light. More importantly, Luminous efficiency was evaluated under a drive current of 150mA, and the highest luminous efficiency, ~95lm/W, was achieved at a mole ratio of Al₂O₃/YAG = 0.65, as shown in Table 2. Notably, the Al₂O₃-Ce:YAG ceramic phosphor has a better luminous efficacy than the transparent Ce:YAG ceramic phosphor under the same test condition. The Al₂O₃ particles play an important role in promoting the luminous efficacy. The Al₂O₃ particle not only changes the propagation of the light in ceramics, but also improves the extraction efficiency of light that may be equivalent to the photonic crystal and surface roughening [17]. The light extraction of the transparent Ce:YAG ceramic phosphor is relatively low in accordance with trapping of light by total internal reflection and the waveguide effect. As seen from Fig. 6(b), the structure of the Al₂O₃-Ce:YAG ceramic makes the propagation direction of the light more variable, thus the extraction efficiency is increased. Obviously, the size and distribution density of Al₂O₃ particles in the YAG matrix have great influence on the luminous efficacy of WLED.

Table 2. WLED luminous efficiency with ceramic samples as the phosphor

View Table | View all tables in this article

 

Fig. 6 The exhibition of the blue light being absorbed and converted by Ce³⁺ (a) and the light propagation (b) in single Ce:YAG matrix and in Ce:YAG matrix with Al₂O₃ particles, the critical angle φc = 33°.

Download Full Size | PDF

4. Conclusions

In summary, we have successfully fabricated Al₂O₃-Ce:YAG ceramics by solid-state reactions. The Al₂O₃ particles sized 2-3μm are homogeneously dispersed in the Ce:YAG matrix whose grain size is 5-15μm. The ceramic sample with the mole ratio of Al₂O₃/YAG = 0.65 has a high luminous efficacy, ~95lm/W. In the Ce:YAG matrix, the existence of the Al₂O₃ particle can change the propagation of the light in ceramics to increase the probability of blue light to be absorbed and converted by Ce³⁺. Furthermore, appropriate amount of Al₂O₃ particles interspersed in Ce:YAG matrix can also improve luminous efficacy of the WLED. The Al₂O₃ particles improve the extraction efficiency of light as the photonic crystal and surface roughening, and meanwhile the fabrication of the Al₂O₃-Ce:YAG ceramic phosphor is easier and with lower cost than the photonic crystal and surface roughening. The result indicates that the investigated Al₂O₃-Ce:YAG ceramic phosphor should be an excellent alternative to the transparent Ce:YAG ceramic phosphor for high-power WLED application. In the next work, we will adjust the distribution density of the Al₂O₃ particle further uniformity and its size more appropriate to improve the luminous efficacy.