Sn–P–F containing glass matrix for the fabrication of phosphor-in-glass for use in high power LEDs

Abstract

In this study, we introduce a low-melting-point (MP) glass ceramic (GlaC) material which can be used to fabricate a phosphor-in-glass (PiG) with Y₃Al₅O₁₂:Ce³⁺ (YAG) phosphor for the realization of high-power white-light-emitting diodes (WLEDs). PiG using a low-MP Sn–P-rich GlaC material, with simple fabrication at 230 °C, can address the drawbacks of the phosphor thermal degradation of currently developed PiGs with a high MP of more than 400 °C, and the yellowing of commercialized phosphor-in-silicon binder (PiSB) materials. In addition, the low-MP PiG can serve as a heat sink due to its good thermal conductivity and good stability compared to those of commercialized PiSB materials. The optical properties of a low-MP PiG-based WLED are a luminous efficacy (LE) of 118 lm W⁻¹, an external quantum efficiency (EQE) of 0.33 and a color-rendering index (CRI) of 68 at 4275 K at an applied current of 350 mA. The simply fabricated and stable PiG-based WLED using the low-MP GlaC material will be a competitive candidate in the WLED lighting market.

---

Introduction

Recently, phosphor-converted white light-emitting diodes (pc-WLEDs) have spurred great interest among most researchers who study lighting issues due to their use in LCD backlighting applications, automobile headlights and general lighting purposes (interior and exterior lighting) owing to their numerous advantages – such as increased longevity, reduced energy use, eco-friendliness, small volume, excellent optical performance capabilities (e.g., a high luminous efficiency (LE), tunable correlated color temperature (CCT), and high color rendering index (CRI)).^1–4 Currently commercialized white pc-LEDs normally use a heat-curable silicone binder with high optical transmittance and a low fabricating temperature as a matrix material to disperse phosphors uniformly onto LED chips while retaining the shape of the LED package. However, the silicone binder used can yellow due to the high heat radiation from the high-power white LED owing to its low glass-transition temperature and the dissociation of the methyl group after operation.^5,6 Eventually, this yellowing problem reduces the transparency of the silicone binder and decreases the optical performance of high-power white pc-LEDs. Therefore, it is necessary to develop alternative materials which are thermally stable and which offer a durable matrix for application to automobile headlights and exterior lighting, where high-power LEDs are typically used. Recently, due to the improved thermal stability and thermal conductivity, transparent glass-ceramic (GlaC) materials were developed as a matrix material to replace silicone binders.^7–9 Therefore, many research groups and companies have suggested many types of novel GlaC materials to make phosphor-in-glass (PiG), in which a certain amount of powder phosphor is dispersed in a transparent GlaC matrix, as thermally stable color converters for pc-LEDs^7–19 (see Table 1).

Table 1 Optical properties of previously-developed PiG materials

Glass ceramic material					LED					Ref.
Composition	Thermal conductivity (W m^−1 K^−1)	Curing condition	Refractive index	Transmittance	LED type	Operating power	LE (lm W^−1)	CIE x	CIE y
Al2O3 SiO2	—	930–965 °C	—	—	—	—	—	—	—	7
SiO2B2O3RO (R = Ba, Zn)	0.5–1.2	750–800 °C	1.57	70–75	High power LED	—	—	—	—	8
Sb2O3 B2O3 TeO2 ZnO Na2O La2O3 BaO	0.71	540–690 °C	1.80	—	High power LED	350 mA	124	0.312	0.333	9
SiO2–Li2O–SrO–Al2O3–K2O–P2O5	1.07	1500 °C → 500 °C	—	—	COB	—	—	—	—	10
Ga2O3	—	1400 °C → 570 °C → 260 °C	—	—	—	350 mA	82	0.321	0.342	11
50PbO–35B2O3–15SiO2	—	500–650 °C	1.8	—	—	—	—	—	—	12
Na2O CaO SiO2	—	800–1000 °C	—	—	LED	1 W	—	—	—	13
TeO2–B2O3–ZnO–Na2O	—	800 °C → 550 °C → 270 °C	1.61–1.87	—	High power LED	350 mA	71	0.375	0.372	14

---

However, the pc-LED, which is realized by using YAG based PiG, showed high coordinate color temperature and a low color rendering index. To solve these problems and improve the color quality, recent reports have demonstrated vision-enhanced PiG by mixing red-colored CaAlSiN₃:Eu²⁺ phosphor, yellow YAG, and inorganic materials that can be converted to a glass-like substance by glassification.^20,21

PiG plates should have two important characteristics: excellent thermal conductivity to diffuse the accumulated heat on the phosphor layer of a high-power pc-LED, and a high refractive index to match that of the phosphors to reduce the light scattering loss of conventional white pc-LEDs.²² Recently, many research groups have conducted experiments to realize low melting point glass frit (under 500 °C) while minimizing the damage to phosphors and thermal degradation.^23,24 In addition, the decreased melting point (MP) of the GlaC matrix has been the subject of research recently, as it is directly coated onto plastic-package or chip on board (COB)-type LEDs and on PiG plates in remote-type pc-LEDs. A lower process temperature is also important to replace the silicone resin with a GlaC material owing to the thermal degradation of the photoluminescence properties of conventional phosphors above 400–500 °C. Accordingly, a low processing temperature and a short processing time are beneficial to maintain the luminescence intensity of the phosphor when fabricating PiG plates.²⁵ However, there have been no reports of a highly transparent GlaC matrix with a MP lower than 400 °C to the best of our knowledge. The optical properties of previously developed PiG materials for WLEDs are summarized in Table 1. In this paper, we summarize and discuss the physical properties of our novel, low-MP fluoride-based GlaC matrix and the optoelectrical down-conversion (DC) performance capabilities of PiG plates with this material on a cup-type white pc-LED package (PiG-on-LED package) and of PiG plates to serve as remote-type plates on COB-type LEDs (PiG on COB) (see Fig. 1). In addition, we describe the method used to fabricate the PiG paste in the LED package and PiG plates with YAG phosphors on COB LEDs using low-temperature co-sintering technology involving fluoride GlaC materials. We also discuss the potential for these innovative low-MP PiG plate-based pc-WLEDs to replace commercial pc-WLEDs.

Fig. 1 Sintering process of phosphor-in-glass and a schematic diagram of phosphor-in-glass on a high-power LED and COB.

Experimental methods

Materials

Glass frit (Sn–P–F based amorphous materials, Yamato Electronics) is synthesized using solid state reaction. The component ratio of glass frit is (100 − x − y) mol% of SnO, x mol% of SnF₂, and y mol% of P₂O₅. The ranges of x and y are 40 to 55, and 20 to 30, respectively. For fabrication of the glass frit the precursor (SnO, SnF₂, and P₂O₅) is mixed uniformly and then melted by using a furnace at 800 °C for 1 hour. The clusters of glass frits are then thermal-quenched and subsequently grinded to a powder. Finally, the powder was sifted to the same size through the sieve. Y₃Al₅O₁₂:Ce³⁺ (YAG, Merck, YGA 585 200), silicone binder (OE-6636 A/B kit, Dow Corning Co.), and blue InGaN LEDs (λ_max = 445 nm, Dongbu LED, Inc.) are used in this study.

Sintering process

Fig. 1 shows the overall scheme of the fabrication process of the PiG plates with the YAG. In the first step, to fabricate the PiGs, appropriate amounts of glass frit and YAG phosphor are mechanically mixed in a bottle. The prepared mixtures are then loaded into an Al round frame. The mixture of phosphor and glass powders is then co-sintered on a hot plate at 230 °C for 0.5–2 minutes. After heating, the sintered PiGs are cooled and easily separated from the Al frame by dropping the plates from the Al frame. Subsequently, the sintered PiGs are post-annealed at 150 °C for one day to eliminate air bubbles and defects inside them.

Fabrication of the silicone binder film

To fabricate the phosphor-in-silicone binder (PiSB) film, silicone binders A and B are mixed at a ratio of 1 : 2. Proper amounts of phosphor are then put into the mixed silicone binder. The glass substrate used for this was then cleaned and treated UV-O₃ for 330 s. Poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS), which is water soluble polymer was spin-coated onto the glass substrate at 1500 rpm to separate glass substrate and phosphor film layer. To obtain ∼800 μm thick films of PiSB, the phosphor binder paste is printed onto the glass substrate within a square-type wall using 3 M tape. The mixture of silicone binder and the YAG phosphor on a glass substrate are sequentially cured at 100, 120, and 150 °C for one hour respectively. The thermally cured films are finally immersed in distilled water and separated from the glass substrate.

Characterization

The transmittance levels of the sintered PiGs and the photoluminescence properties of the phosphor after the heat treatment were measured using a Xe lamp and a spectrophotometer (Darsa, PSI Trading Co., Ltd.) respectively. The structural properties of the fabricated PiGs were analyzed using X-ray diffraction (XRD) with Cu Kα radiation (D-max 2500, Rigaku). Surface and cross-sectional images of the PiGs were obtained by scanning electron microscopy (SEM, JSM-7401F, JEOL Ltd.). The components of the PiGs were then measured with an energy dispersive spectrometer (EDS). The densities of the PiGs and the silicone binder film were measured using an electronic densimeter (MD-300S, Alfa Mirage). The specific heat and thermal diffusivity were determined through a plastics-specific differential scanning calorimetry method (DSC Q20, TA) and with thermal diffusivity measurements (LFA 447 NanoFlash, NETZSCH), respectively. The thermal conductivity was calculated using the density, specific heat, and thermal diffusivity. The optical properties of the PiG plate- and PiSB film-based DC-WLEDs were measured in an integrated sphere using a spectrophotometer.

Result and discussion

Fig. 2a indicates that the transmittance of the Glac plate between 380 and 780 nm was in the range of 71–80%. The inset shows a photograph of a Glac plate with a thickness of 0.8 mm prepared by sintering Glac powder at 230 °C for 45 s. Fig. 2b and c show that transmittance values of PiGs with various weights of YAG phosphor and Glac materials mixture and weight percentages of the YAG phosphor in the PiGs. Spectra at 460 nm indicate the typical 4f → 5d transition of Ce³⁺. The transparency and low-temperature MP of the Glac materials indicate the possible applicability of the matrix materials for PiGs in pc-WLEDs. Table 2 summarizes the measured optical and physical properties of this Glac material in comparison with related parameters of the silicone resin used for dispersing phosphors in conventional pc-LEDs.

Fig. 2 Transmittance of (a) the bare glass plate, (b) the PiG by weight, and (c) the PiG by weight percentage. All samples were fabricated at 230 °C for 45–90 s.

Table 2 Properties of glass and silicone resin

Materials properties	Hardness (Mpa)	Refractive index	Glass transition temperature (°C)	Coefficient of linear expansion (ppm per °C)	Transmittance (T%)
Glac material	>1000	1.60–1.90	100–140	15–18	71–80
Silicone resin	1–100	1.30–1.55	10–30	20–50	85–92

---

The refractive index (n = 1.60–1.90) of the low-MP Glac material is higher than that of the silicone resin (n = 1.30–1.55) and approaches that of the YAG phosphor (n = 1.84, at 550 nm). Thus, the reduced difference in the refractive index between the Glac and the phosphor can reduce the light-scattering loss of the co-sintered PiG plate. Table 2 also indicates that the mechanical properties of the Glac plate are superior to those of the silicone resin. The excellent mechanical properties of the Glac plate provide evidence that using it as an outer packaging material for pc-LEDs is feasible. Furthermore, the higher thermal conductivity and lower thermal expansion coefficient of low-MP Glac materials also provide benefits for rapid heat transfer to air in high-power pc-WLEDs. This low-MP Glac material is a promising candidate as a packaging material for dispersing phosphors in pc-WLEDs owing to its better thermal and mechanical properties and similar optical properties relative to those of silicone resins.

As a result of the co-sintering of the YAG phosphor and the Glac matrix, the PiG composite plates were well fabricated as remote-type phosphor plates for high power LEDs. Fig. 3 shows the photoluminescence (PL) and PL excitation (PLE) spectra of a YAG PiG plate and the corresponding YAG powder phosphor dispersed on the silicone binder film. The two samples showed similar PL and PLE spectra, as indicated by the powder phosphors in this experiment and in previous publications.^10,22 Both broad and bright yellow emissions are centered at ∼545 nm (Ce³⁺: 5d → 4f) under excitation at 450 nm. The temperature-dependence of the PL intensity of the powder phosphor indicates that the low co-sintering temperature does not detrimentally affect the PL efficiency of the PiG material during the co-sintering process (see ESI Fig. S1†). In the PLE spectra (340, 460 nm peaks: Ce³⁺: 4f → 5d), the similarity between the two samples confirms the negligible heat effect of the low co-sintering temperature, despite the fact that the decreased 340 nm PLE intensity is due to the absorption of the Glac matrix in the short UV wavelength range. The effects of the co-sintering temperature on the PL performance capabilities of YAG PiGs were also investigated (see Fig. S2†) as a function of the sintering temperature. The EL intensity of blue light at 454 nm as measured with a bare glass plate rises with an increase in the co-sintering temperature to 230 °C, though it is clearly degraded above 230 °C, indicating that the PiG sintered at a temperature exceeding 230 °C leads to reduced transparency of Glac plates. Thus, the external quantum efficiency (EQE) of the PiG material reached its maximum value at 230 °C and decreased above 230 °C (see Fig. S2†). The EQE was calculated by following eqn (1)

Fig. 3 PL and PLE spectra of (a) PiG plate, and (b) PiSB film.

In contrast to results pertaining to the co-sintering time in numerous previous reports, our fluoride-based Glac materials are very quickly melted and co-sintered with the YAG phosphor after heating for less than 1 min. After careful and systematic evaluations of the measured transparency levels of the Glac and PiG plates and the PL intensity as well as the EQE of a YAG PiG placed on a blue LED package, the PiG with 3 wt% YAG phosphor which has a CCT similar to that of PiSB, sintered at 230 °C for 45 s was regarded as the optimum condition for the fabrication of PiG color converters (see Fig. S3†). In order to reduce the possibility of impaired transparency due to the existence of a large number of tiny gas bubbles in the PiG when the co-sintering temperature is too low and/or the co-sintering time is too short, the PiG plates were post-annealed at 150 °C for 1 day. Fig. 4a and c shows the down-converted EL (DC-EL) spectra and Fig. 4b and d shows the CIE color coordinates after the post-annealing of the YAG-based PiG plates. As expected, the DC-EL efficiency of the post-annealed sample was enhanced by ∼5% (see Fig. S4 and S5†); this enhancement is due to the reduced scattering loss of the decreased bubbles and the increased the density of PiG plate. The optical properties of the post-annealed YAG-based PiG plate-capped DC-WLEDs are summarized in Table 3. Fig. 5a and b show SEM images of PiG plates containing 3 wt% YAG phosphor before and after the post-annealing process. The microstructure of the post-annealed PiG substrate confirms that YAG phosphor particles 10–15 μm in size are uniformly dispersed in the denser Glac matrix. A SEM-EDS element analysis also indicated that the YAG phosphor particles are clearly distinguished from the Glac matrix. As shown in Fig. 5c, the Sn- or P-rich region presents the glass matrix and the Y- or Al-rich region indicates the phosphor particles. The uniform distribution of the phosphor particles and the reduced refractive index difference resulted in low light scattering loss and uniformly emitted light.⁷ However, a homogeneous distribution of phosphor particles is too difficult to realize in conventional silicone paste owing to the inhomogeneous distribution of the phosphor in the silicone resin. XRD patterns of a bare glass plate, the YAG phosphor powder, and a fabricated PiG are shown in Fig. 6 for comparison. The low-MP bare glass plate has two broad amorphous peaks. The XRD patterns of a PiG with a broad peak and showing several sharp peaks as well indicate that the YAG powder and glass frit are well mixed and sintered within the crystal scale. To test the possibility of preparing PiG plates for application to DC-WLEDs, a WLED package is constructed by placing YAG PiG plates on a blue cup-type LED package and a COB LED, as shown in Fig. 1. For comparison, the PiSB film is used as a reference sample. Fig. 7a and c show the white DC-EL spectrum of post-annealed PiG plates on a blue LED package while varying the PiG weight on a COB LED or a high-power cup-type LED package. The blue intensity levels of the PiG-based LED decrease and the yellow intensity levels increase with an increase of the PiG weight. The variations of the color coordinates of the PiG on both the LED and the COB LED are shown as a function of the PiG weight in Fig. 7b and d, respectively. The CIE color temperatures of the DC-WLED covered with the PiG plate are easily tunable by a simple change of the PiG wt% in the plates, as predicted in previous publications. The LE of the PiG-based WLED after post-annealing the PiG material shows good values between 108 and 125 lm W⁻¹. The DC-EL properties of both the COB-type LEDs and the cup-type LEDs were measured at a rated voltage of 36 V and a rated current of 350 mA, respectively. The EQE values of PiG-based cup-type LEDs were found to be somewhat higher than those of the PiG-based COB-type LEDs. It is surmised that this EQE difference is due to the higher working temperature of the COB-type LEDs. The detailed optical properties of all PiG-based LEDs are summarized as a function of the thickness of the PiG plates in Table 4. Thicker plates reduce the blue spectrum and move the CIE color coordinates from the white region to the yellow region. The thicker films also reduce the CRI values owing to the misbalance between blue and yellow light to produce the appropriate white color. Therefore, it is necessary to add red or orange phosphors to the YAG-based PiG to improve the color quality of WLEDs, as reported in many publications related to high-CRI LEDs.^10,12,16

Fig. 4 (a) and (b) EL spectra and color coordinates of PIG after post-annealing on COB at 36 V, and (c) and (d) EL spectra and color coordinates of PIG after post-annealing on high power LED at 350 mA. PiGs were sintered at 230 °C for 45 s.

Table 3 EL data of PIG sintered at 230 °C for 45 s (a) on COB at 36 V (b) on high power LED at 350 mA

	CCT	CIE x	CIE y	LE (lm W^−1)	LER (lm Wopt^−1)	EQE	CRI
COB	4370	0.369	0.394	110	363	0.303	64
High-power LED	4275	0.374	0.399	118	358	0.330	68

---

Fig. 5 SEM images of PiG (a) before annealing, (b) after post-annealing. The insets show a cross section image of each PiG sample. The thickness of PiG is (a) 820 μm and, (b) 835 μm respectively. The sintering temperature of PiG was 230 °C and the post annealing temperature was 150 °C. (c) EDS data of PiG. Each part shows the GlaC material and YAG phosphor.

Fig. 6 XRD patterns of bare glass plate, YAG powder, and 5 wt% PiG.

Fig. 7 EL spectra and color coordinates of PiG based DC-WLED (a) and (b) on COB (c) and (d) on high power LED with various PiG weight. The insets of (b) and (d) show photographs of the emission light from integrated sphere with an increase in the weight of PiG.

Table 4 Thickness and optical properties of PIGs on COB at 36 V and on high power LED at 350 mA. Sintering temperature was 230 °C

	g	Thickness (μm)	Color coordinates	LE (lm W^−1)	CCT (K)	EQE	CRI
COB	0.8	600	(0.333, 0.334)	108	5415	0.335	70
	1.0	800	(0.369, 0.394)	111	4370	0.303	64
	1.2	1000	(0.383, 0.416)	110	4135	0.292	62
	1.4	1200	(0.402, 0.444)	111	3885	0.279	61
	1.6	1400	(0.412, 0.460)	110	3790	0.270	60
LED	0.8	600	(0.326, 0.327)	111	5740	0.358	78
	1.0	800	(0.374, 0.399)	118	4275	0.330	68
	1.2	1000	(0.384, 0.414)	120	4110	0.325	67
	1.4	1200	(0.408, 0.451)	124	3810	0.318	64
	1.6	1400	(0.417, 0.465)	125	3720	0.314	63

---

Higher thermal conductivity is more efficient for releasing the heat of the phosphor layer of high-power cup-type and COB-type LEDs. Here, PiSB, which is widely used as an encapsulant, is also used as a reference sample for comparison. Fig. 8 shows the thermal diffusivity curves of both the PiG plates and PiSB films with similar thicknesses. The thermal conductivity is calculated using the measured density, specific heat and thermal diffusivity, as noted in previous publications. Thermal conductivity was calculated by following eqn (2)

Thermal conductivity (W m⁻¹ K⁻¹) = density (g cm⁻³) × specific heat (J g⁻¹ K⁻¹) × thermal diffusivity (mm² s⁻¹) (2)

Fig. 8 Results of thermal diffusivity by the flash method. (a) PiSB film (b) PiG plate.

After calculating from the thermal diffusivity curves, Table 5 summarizes the thermal characteristics of the PiG plate and PiSB. This table indicates that the thermal conductivity of the YAG PiG is 1.259 W m⁻¹ K⁻¹, which is 1.62 times higher than that of the YAG PiSB film.

Table 5 Thermal conductivity PiSB film and PiG plate

	Density (g cm^−3)	Specific heat (J g^−1 K^−1)	Thermal diffusivity (mm^2 s^−1)	Thermal conductivity (W m^−1 K^−1)
PiSB film	1.255	1.177	0.527	0.778
PiG plate	3.795	0.676	0.491	1.259

---

The current-dependences of the PiG plates and the PiSB films on the cup-type LEDs and COB LEDs are shown in Fig. 9a. These figures indicate that the LE of both the PiG plates and PiSB films on high-power WLEDs decrease with an increase in the applied current from 50 mA to 700 mA. In the high current region above 500 mA, the normalized LEs of the PiSB films show lower values than those of the PiG plates. Eventually, the normalized LE of the PiG plate decreases to 52.3% while that of the PISB film decreases to 47.5% at 600 mA. It can be surmised that this 10% difference is due to the difference in the thermal conductivity values between the PiG and PiSB samples. Fig. 9b also shows the temperature-dependence of the normalized LE of PiG plates and PiSB films on cup-type high-power LEDs. Over a temperature of 100 °C, the normalized LE values of PiG plates slowly decrease with an increase in the temperature to 160 °C, whereas those of the PiSB films decrease more rapidly as compared to the PiG plates. It can therefore be considered that the significant difference in the normalized LE between the PiG plates (85.3%) and the PiSB films (74.5%) at 160 °C is due to both the detrimental effects of the yellowing phenomenon of the silicone binder heated a temperature exceeding 140 °C and the low thermal conductivity of the PiSB films. Fig. 9c shows the effect of varying the applied voltage on the normalized LEs of PiG plates and PiSB films capped onto COB-type WLEDs.

Fig. 9 Various dependence of PIG and PiSB films. (a) Current-dependence on high power LED, (b) temperature-dependence on high-power LED, and (c) voltage-dependence on COB.

Better voltage stability of the normalized LEs is obtained from the PiG-capped COB-type WLEDs. Due to the similar origins to the better current- and temperature-dependencies of the PiG-capped cup-type LED, the PiG plates show better voltage stability than the PiSB film-capped WLEDs.

The shifts of the CIE color coordinates of PiG plate- and PiSB film-based DC-WLEDs are compared in Fig. 10a–c. The CIE color coordinates of the PiG plates and PiSB films shifted to a bluish color with an increase in the applied current, temperature and applied voltage.

Fig. 10 CIE shift of PiG plate and PiSB film based DC-WLED with (a) the applied current on a high power LED, (b) applied temperature on high power LED, (c) applied voltage on COB. Changes of the CCT of PiG plate and PiSB film based DC-WLED with (d) the applied current on a high power LED, (e) applied temperature on high-power LED, and (f) applied voltage on COB.

These figures indicate that the changes in the CIE color coordinates of the PiSB films are much greater than those of the PiG plates with an increase in the current, temperature, and voltage variables, irrespective of whether a cup-type or a COB-type LED is used. The difference in the CIE color shifts between the PiG plates and the PiSB films is clearly distinct owing to the better current, thermal, and voltage stability levels of the PiG plates. As noted above, a PiG plate is more stable at a higher temperature owing to its higher thermal conductivity and lower reactivity with air than PiSB film. Fig. 10d–f also clearly indicate that the shifts of the CCT of the PiG plates are less than those of the PiSB films with increases in the current, temperature and voltage, irrespective of the package type of WLED used. All figures of the LE, CIE color temperature, and CCT dependences of the PiG plates collectively confirm that the PiG plates have better thermal, current, and voltage stability levels than PiSB films. PiG shows better stability despite its lower LE value compared to that of PiSB. In addition, we conducted a long-term stability test of PiG on a high power LED. The luminous efficacy of PiG based DC-WLED maintained constantly up to 100 hours. Also, the structure properties of PiG plate unchanged (see Fig. S6†).

Conclusions

In this study, low-melting-temperature, transparent, and fluoride-based Glacs were prepared using Sn–P–F-based amorphous powders. These low-MP Glacs are considered to be promising candidates as matrix materials for PiG plates in pc-WLEDs. We analyzed the optical and electrical properties and the morphological structures of post-annealed PiG plates to assess their suitability as color-conversion materials which may replace phosphors dispersed in silicone binder. We then characterized the optoelectrical performances of PiG-capped LEDs as a function of various operational parameters. The resultant PiG-based white DC-WLEDs, i.e., a YAG PiG plate on a cup-type LED, showed a LE value of 118 lm W⁻¹ at 4275 K, an EQE of 0.330, and a CRI of 68 at an applied current of 350 mA fabricated with 3 wt% YAG phosphor at 230 °C for 45 s. The post-annealed and co-sintered YAG PiG plate displayed excellent heat dissipation performance, in contrast with conventional PiSB films. The lower sintering temperature can be beneficial when attempting to decrease the thermal degradation of the photoluminescence properties of conventional phosphors. The thermal, current and voltage stability levels of PiG-capped WLEDs are much better than those of PiSB-capped WLEDs owing to the better thermal conductivity and lower reactivity of PiG plates. Accordingly, it is likely that PiG-based WLEDs will advance the development of highly stable WLED forms of lighting in various lighting markets, such as LCD backlighting, automobile headlights and general lighting applications.

Acknowledgements

This work was supported by a grant of the National Research Foundation of Korea (University-Institute cooperation program) and by the Energy Technology Development Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant (No. 20143030011530), and the Ministry of Science, ICT&Future Planning (MSIP) (No. 2016R1A5A1012966), funded by the Korean Government.

Notes and references

1. P. Pust, V. Weiler, C. Hecht, A. Tücks, A. S. Wochnik, A.-K. Henß, D. Wiechert, C. Scheu, P. J. Schmidt and W. Schnick, Nat. Mater., 2014, 13, 891–896.
2. S. Ye, F. Xiao, Y. X. Pan, Y. Y. Ma and Q. Y. Zhang, Mater. Sci. Eng., R, 2010, 71, 1–34.
3. S. Pimputkar, J. S. Speck, S. P. DenBaars and S. Nakamura, Nat. Photonics, 2009, 3, 180–182.
4. N. Kimura, K. Sakuma, S. Hirafune, K. Asano, N. Hirosaki and R.-J. Xie, Appl. Phys. Lett., 2007, 90, 051109.
5. T. Yanagisawa and T. Kojima, J. Lumin., 2005, 114, 39–42.
6. C. Tsai, W. Cheng, J. Chang, L. Chen, J. Chen, Y. Hsu and W. Cheng, J. Disp. Technol., 2013, 9, 427–432.
7. L. Wang, L. Mei, G. He, J. Li and L. Xu, J. Am. Ceram. Soc., 2011, 94, 3800–3803.
8. Y. K. Lee, J. S. Lee, J. Heo, W. B. Im and W. J. Chung, Opt. Lett., 2012, 37, 3276–3278.
9. R. Zhang, H. Lin, Y. Yu, D. Chen, J. Xu and Y. A. Wang, Laser Photonics Rev., 2014, 8, 158–164.
10. X. Zhang, L. Huang, F. Pan, M. Wu, J. Wnag, Y. Chen and Q. Su, ACS Appl. Mater. Interfaces, 2014, 6, 2709–2717.
11. H. Lin, B. Wang, J. Xu, R. Zhang, H. Chen, Y. Yu and Y. Wang, ACS Appl. Mater. Interfaces, 2014, 6, 21264–21269.
12. S. Yi, W. J. Chung and J. Heo, J. Am. Ceram. Soc., 2014, 97, 342–345.
13. A. Herrmann, C. Russel and P. Pachler, Opt. Mater. Express, 2015, 5, 2193–2200.
14. H. Chen, H. Lin, J. Xu, B. Wang, Z. Lin, J. Zhou and Y. Wang, J. Mater. Chem. C, 2015, 3, 8080–8089.
15. J. S. Lee, P. Arunkumar, S. Kim, I. J. Lee, H. Lee and W. B. Im, Opt. Lett., 2014, 39, 762–765.
16. Y. Zhou, D. Chen, W. Tian and Z. Ji, J. Am. Ceram. Soc., 2015, 98, 2445–2450.
17. S. Kim, H. Yie, S. Choi, A. Sung and H. Kim, Opt. Express, 2015, 23, A1499–A1511.
18. J. Huang, X. Liang, W. Xiang, M. Gong, G. Gu, J. Zhong and D. Chen, Mater. Lett., 2015, 151, 31–34.
19. Z. Lin, H. Lin, J. Xu, F. Huang, H. Chen, B. Wang and Y. Wang, J. Eur. Ceram. Soc., 2016, 36, 1723–1729.
20. D. Q. Chen, Y. Zhou, W. Xu, J. S. Zhong, Z. G. Ji and W. D. Xiang, J. Mater. Chem. C, 2016, 4, 1704–1712.
21. H. Park, Y. K. Lee, W. B. Im, J. Heo and W. J. Chung, Opt. Mater., 2015, 41, 67–70.
22. B. K. Park, H. K. Park, J. H. Oh, J. R. Oh and Y. R. Do, J. Electrochem. Soc., 2012, 159, 96–106.
23. J. Huang, X. Hu, J. Shen, D. Wu, C. Yin, R. Xiang, C. Yang, X. Liang and W. Xiang, CrystEngComm, 2015, 17, 7079–7085.
24. X. Zhang, J. Yu, J. Wang, C. Zhu, J. Zhang, R. Zou, B. Lei, Y. Liu and M. Wu, ACS Appl. Mater. Interfaces, 2015, 7, 28122–28127.
25. Y. K. Lee, Y. H. Kim, J. Heo, W. B. Im and W. J. Chung, Opt. Lett., 2014, 39, 4084–4087.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra23294a

‡ These authors contributed equally to this work.

---