Warm-white light-emitting diode with high color rendering index fabricated by combining trichromatic InGaN emitter with single red phosphor Jinn-Kong Sheu,1,2,* Fu-Bang Chen,1,3 Yen-Chin Wang,1,3 Chih-Chiang Chang,4 ShihHsien Huang,3 Chun-Nan Liu,1 and Ming-Lun Lee5 1 Department of Photonics, National Cheng Kung University, Tainan City, 70101, Taiwan Advanced Optoelectronic Technology Center and Research Center for Energy Technology & Strategy, National Cheng Kung University, Tainan City, 70101, Taiwan 3 High Power Opto Inc., Taichung Science Park, Taichung City, 407, Taiwan 4 Excellence Optoelectronics Inc., Hsinchu Science Park, Hsinchu, 35053, Taiwan 5 Department of Electro-Optical Engineering, Southern Taiwan University of Science and Technology, Tainan City, 71005, Taiwan * [email protected] 2
Abstract: We present a trichromatic GaN-based light-emitting diode (LED) that emits near-ultraviolet (n-UV) blue and green peaks combined with red phosphor to generate white light with a low correlated color temperature (CCT) and high color rendering index (CRI). The LED structure, blue and green unipolar InGaN/GaN multiple quantum wells (MQWs) stacked with a top p-i-n structure containing an InGaN/GaN MQW emitting n-UV light, was grown epitaxially on a single substrate. The trichromatic LED chips feature a vertical conduction structure on a silicon substrate fabricated through wafer bonding and laser lift-off techniques. The blue and green InGaN/GaN MQWs were pumped with n-UV light to re-emit low-energy photons when the LEDs were electrically driven with a forward current. The emission spectrum included three peaks at approximately 405, 468, and 537 nm. Furthermore, the trichromatic LED chips were combined with red phosphor to generate white light with a CCT and CRI of approximately 2900 and 92, respectively. ©2015 Optical Society of America OCIS codes: (230.3670) Light-emitting diodes; (230.0250) Optoelectronics; (230.5590) Quantum-well, -wire and -dot devices.
References and links 1. 2. 3. 4. 5. 6.
E. F. Schubert and J. K. Kim, “Solid-state light sources getting smart,” Science 308(5726), 1274–1278 (2005). G. Verzellesi, D. Saguatti, M. Meneghini, F. Bertazzi, M. Goano, G. Meneghesso, and E. Zanoni, “Efficiency droop in InGaN/GaN blue light-emitting diodes: physical mechanisms and remedies,” J. Appl. Phys. 114(7), 071101 (2013). J. H. Oh, J. R. Oh, H. K. Park, Y. G. Sung, and Y. R. Do, “New paradigm of multi-chip white LEDs: combination of an InGaN blue LED and full down-converted phosphor-converted LEDs,” Opt. Express 19(S3 Suppl 3), A270–A279 (2011). K. J. Chen, H. T. Kuo, H. C. Chen, M. H. Shih, C. H. Wang, S. H. Chien, S. H. Chiu, C. C. Lin, C. J. Pan, and H. C. Kuo, “High thermal stability of correlated color temperature using current compensation in hybrid warm white high-voltage LEDs,” Opt. Express 21(S2 Suppl 2), A201–A207 (2013). C. C. Lin and R. S. Liu, “Advances in phosphors for light-emitting diodes,” J. Phys. Chem. Lett. 2(11), 1268– 1277 (2011). J. Y. Tsao, M. H. Crawford, M. E. Coltrin, A. J. Fischer, D. D. Koleske, G. S. Subramania, G. T. Wang, J. J. Wierer, and R. F. Karlicek, Jr., “Toward smart and ultra-efficient solid-state lighting,” Adv. Opt. Mater. 2(9), 1– 28 (2014).
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A232
7. 8. 9. 10. 11. 12. 13. 14.
J. M. Phillips, M. E. Coltrin, M. H. Crawford, A. J. Fischer, M. R. Krames, R. Mueller-Mach, G. O. Mueller, Y. Ohno, L. E. S. Rohwer, J. A. Simmons, and J. Y. Tsao, “Research challenges to ultra-efficient inorganic solidstate lighting,” Laser Photon. Rev. 1(4), 307–333 (2007). J. K. Sheu, F. B. Chen, S. H. Wu, M. L. Lee, P. C. Chen, and Y. H. Yeh, “Vertical InGaN-based green-band solar cells operating under high solar concentration up to 300 suns,” Opt. Express 22(S5 Suppl 5), A1222–A1228 (2014). J. K. Sheu, S. J. Chang, C. H. Kuo, Y. K. Su, L. W. Wu, Y. C. Lin, W. C. Lai, J. M. Tsai, G. C. Chi, and R. K. Wu, “White-light emission from near UV InGaN-GaN LED chip precoated with blue/green/red phosphors,” IEEE Photon. Technol. Lett. 15(1), 18–20 (2003). D. Schiavon, M. Binder, A. Loeffler, and M. Peter, “Optically pumped GaInN/GaN multiple quantum wells for the realization of efficient green light-emitting devices,” Appl. Phys. Lett. 102(11), 113509 (2013) (and references therein). Y. Chen, M. Gong, G. Wang, and Q. Su, “High efficient and low color-temperature white light-emitting diodes with Tb3 Al5O12: Ce3+ phosphor,” Appl. Phys. Lett. 91(7), 071117 (2007). H. S. Jang, W. B. Im, D. C. Lee, D. Y. Jeon, and S. S. Kim, “Enhancement of red spectral emission intensity of Y3Al5O12:Ce3+ phosphor via Pr co-doping and Tb substitution for the application to white LEDs,” J. Lumin. 126(2), 371–377 (2007). S. Nizamoglu, T. Erdem, X. W. Sun, and H. V. Demir, “Warm-white light-emitting diodes integrated with colloidal quantum dots for high luminous efficacy and color rendering,” Opt. Lett. 35(20), 3372–3374 (2010). J. R. Chen, Y. C. Wu, S. C. Ling, T. S. Ko, T. C. Lu, H. C. Kuo, Y. K. Kuo, and S.-C. Wang, “Investigation of wavelength-dependent efficiency droop in InGaN light-emitting diodes,” Appl. Phys. B 98(4), 779–789 (2010).
1. Introduction White light-emitting diodes (WLEDs) have been recognized as a promising next-generation general lighting source in response to the issues of energy shortages and environmental degradation [1]. Commercial WLEDs are produced mainly in two ways. The first selectively combines red, green, and blue (RGB) LEDs. However, this is costly, and the overall light outputs of RGB LEDs degrade at different rates. As a result, the color is eventually unbalanced. The most popular method of producing WLEDs is by combining GaN-based blue LEDs and yellow phosphor. Another possible technique of white light generation involves the combination of deep ultraviolet (UV) LEDs and RGB phosphors that act as fluorescent tubes. Nonetheless, the development of deep UV-pumped WLEDs remains in its infancy because of the limited efficiency of deep UV LEDs. The efficiency of GaN-based LEDs depends strongly on the indium content in the active layer. InGaN LEDs exhibit high external quantum efficiencies (EQEs) in the violet and blue regions. However, this efficiency is significantly lower in green emission devices, which have greater indium content in the InGaN active region. This efficiency loss can be attributed to a number of material issues, including highdensity dislocation and strong polarization fields [2]. Commercial WLEDs are mainly produced through the combination of a single blue (λ = 440–460 nm) LED chip with yellow phosphor (λ ~560 nm), such as a cerium-doped yttrium aluminum garnet (YAG:Ce), in consideration of cost-performance merits. Their efficiency has far exceeded that of traditional incandescent bulbs. Dichromatic WLEDs with a color rendering index (CRI) of approximately 70–80 and a correlated color temperature (CCT) of 4000–8000 K are suitable for outdoor applications [1]; however, these LEDs are inadequate for indoor illumination applications that generally require CRIs higher than 80 [3]. To improve color characteristics, a second red (or orange) LED chip or a red-emitting phosphor has been integrated to generate a trichromatic WLED [4,5]. Similarly, the CRI can be boosted to a maximum of 85 through the multichip approach, which uses LED chips alone to generate white light with color components at red, green, and blue wavelengths [6]. Another multichip approach utilizes four LED chips that emit red, yellow, green, and blue (RYGB) wavelengths to establish a tetrachromatic WLED that can achieve a CRI higher than 95 [7]. This study introduces a trichromatic GaN-based LED that emits near-UV (n-UV), blue, and green peaks combined with red phosphor to generate white light with a low CCT and high CRI. In the spectrum of this trichromatic LED, the blue and green peaks originate from photon recycling using n-UV light excitation, which is generated by electrical pumping, to optically pump two separate multiple quantum wells (MQWs). Blue and green unipolar InGaN/GaN MQW structures are
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A233
grown on the same substrate, which is stacked with a top p-i-n structure containing an n-UVemitting InGaN/GaN MQW. The blue and green InGaN/GaN MQWs are pumped with n-UV light to re-emit low-energy photons when the n-UV MQW is electrically driven by a forward current. The resulting emission spectrum shows three peaks at approximately 405, 468, and 537 nm. The trichromatic LED chips are then combined with red phosphor to generate white light. The trichromatic GaN-based LEDs have a vertical conduction structure on a silicon substrate that is fabricated using wafer bonding and laser lift-off (LLO) techniques to alleviate the thermal effect on performance degradation when the devices operate at high input power [8]. The fabrication procedures and characterizations of the experimental LEDs are detailed in the following sections. 2. Experiments The InGaN-based LEDs used in this study were grown on c-face sapphire substrates using a metalorganic vapor-phase epitaxy reactor. The layer structure consisted of a 30-nm-thick lowtemperature GaN nucleation layer and 3-µm-thick undoped GaN (u-GaN) with a chamber pressure of 500 Torr. A 16-pair green InGaN/GaN MQW containing a 2.5-nm-thick In0.28Ga0.72N well separated by a 9-nm-thick GaN barrier in each pair was also grown in sequence with a 16-pair blue InGaN/GaN MQW containing a 2-nm-thick In0.15Ga0.85N well separated by a 8-nm-thick GaN barrier. The green and blue MQWs emitted light wavelengths of approximately 535 nm and 465 nm, respectively, as determined by photoluminescence (PL) measurements. A p–i–n heterostructure was then grown sequentially on the green and blue MQWs. The heterostructure consisted of a 4-µm-thick Si-doped n+-GaN layer (n ~3 × 1018 cm−3), a 12-pair MQW that contained twelve 3-nm-thick In0.05Ga0.95N wells separated by 12-nm-thick Al0.15Ga0.85N barriers, and a 200-nm-thick Mg-doped p-GaN layer (p ~5 × 1017 cm−3). The typical PL peak wavelength, as obtained from the In0.05Ga0.95N/Al0.15Ga0.85N MQW, was approximately 405 nm, and this MQW was denoted as an n-UV violet MQW. Figure 1(a) shows the schematic of the layer structure of the aforementioned epitaxial heterostructure. A bilayer Ni/Ag metal (1/200 nm) was deposited onto the top p-GaN layer following epitaxial growth to serve as a reflector/ohmic contact layer [8]. Once the reflector layer was formed, a barrier layer consisting of a 200-nm-thick TiW layer and a 50-nm-thick Pt layer was configured to alleviate the diffusion of Ag from the reflector layer to the bonding layer. This barrier layer was deposited between the reflector and the bonding layers. A 3-µmthick In layer was used as a bonding layer in the proposed vertical InGaN/GaN/Si photon recycling LEDs (PRLEDs). In addition, the Si substrates were dipped into a buffered oxide etch solution for 60 s to remove the native silicon dioxide; a bilayer Ti/Au (20/1500 nm) metal was then deposited onto the surfaces of the Si substrates to serve as ohmic contact and bonding layers [8]. These Si wafers acted as receptors during the wafer bonding process. The sapphire substrate was removed by LLO to expose the u-GaN layer after the wafer bonding process. The InxGa1−xN/GaN-based heteroepitaxial layers were thus transferred to the Si substrate along with the u-GaN top layer. The samples were then treated in a potassium hydroxide solution (concentration of 3 M) at an elevated temperature of 60°C to texture the uGaN layer for enhancing light extraction efficiency. The n+-GaN layer was exposed by selective-area dry etching after surface texturing. Ti/Al/Ni/Au (20/30/150/2000 nm) metal layers were then deposited onto this exposed n+-GaN layer to form n-type ohmic contacts (cathode electrodes) on the wafers. Finally, the Si substrates were thinned to 150 µm and coated with Ti/Au (50/500 nm) metals to serve as the rear ohmic contact layer.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A234
Fig. 1. (a) Layer structure of the PRLED epitaxial heterostructure. (b) Typical top-view image of the proposed vertical InGaN/GaN/Si PRLED chip. (c) Schematic cross-section layer structure of the vertical PRLED chip.
Figure 1(b) shows a typical top-view image of the proposed vertical InGaN/GaN/Si PRLEDs as obtained through an optical microscope. The dashed line indicates the cross section that corresponds to a schematic of the layer structure of the vertical LEDs, as shown in Fig. 1(c). All of the experimental PRLED chips used in the present study had an area of 1.25 × 1.25 mm2. To generate the WLEDs, the trichromatic PRLED chips were coated with red phosphor that was mixed with resin. The red phosphor was a Eu-doped (SrCa)AlSiN3 powder with an average particle size of 10 μm. The room-temperature current-voltage (I—V) characteristics of the experimental LEDs were measured using a HP-4156C semiconductor parameter analyzer. The light output power-current (L—I) of the LEDs was measured using a calibrated integrating sphere that was combined with a Keithly 2400 source meter. 3. Results and discussions Figure 2(a) depicts the typical I—V characteristics of the vertical InGaN/GaN/Si PRLEDs and the n-UV LEDs without the stacked optical converters. The forward I—V curves of the LEDs do not differ significantly. In principle, a fraction of pump light, that is, the n-UV light emission generated by direct current injection, is absorbed by the optical converters (i.e., the stacked blue and green InGaN quantum wells). This fraction increases with the total thickness of the converters. However, the lattice mismatch between GaN and InxGa1−xN limits the total thickness of the InxGa1−xN layer in the optical converters of the PRLED structure, particularly when the In content is increased. In other words, the performance of the p–i–n GaN/InxGa1−xN electrical n-UV emitter declines if the total thickness and/or In content of the optical converters are too high, because the n-UV emitter is grown on the optical converters, as indicated in Fig. 1(a). Considering the I—V characteristics, Fig. 2(a) suggests that the epitaxial growth of the p–i–n heterostructure on the optical converters (blue and green MQWs) did not significantly degrade in material quality.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A235
Fig. 2. (a) Typical I-V characteristics of vertical InGaN/GaN/Si PRLEDs and n-UV LEDs without the stacked optical converters. (b) Typical EL spectra of the vertical InGaN/GaN/Si PRLEDs and n-UV LEDs without the stacked optical converters, under an injection current of 350 mA.
Figure 2(b) displays the typical electroluminescence (EL) spectra obtained from vertical InGaN/GaN/Si PRLEDs and the n-UV LEDs without stacked optical converters under an injection current of 350 mA. As expected, the n-UV MQW exhibited a peak at 405 nm because of carrier injection into the In0.05Ga0.95N/Al0.1Ga0.9N MQW and the recombination therein. Aside from this n-UV peak, the PRLEDs displayed two additional peaks at 468 nm and 537 nm, which originated from the optical pumping with current-injected n-UV light. This pumping excited the stacked blue and green MQWs. Nonetheless, the phosphorconverted emissions exhibited broader spectra than the light emissions of the direct currentinjected LEDs or the photon-recycling semiconductors [10], although n-UV LEDs have been shown to excite phosphors through down conversion to generate distinct RGB emission peaks [9]. Luminous efficacy is limited in traditional white light sources based on emissions with broad spectra, because the sensitivity of the human eye declines rapidly as one approaches infrared and UV wavelengths. In addition, the human eye possesses only three types of colorsensitive receptors or cones. Therefore, high-efficiency WLEDs can be produced by generating light with three distinct colors as with the multichip approach, which produces white light using RGB LEDs [1]. Moreover, the assembly of InGaN-based blue and green LEDs and of the AlGaInP-based red LED can enhance luminous efficiency more than other solutions can, although LED performance to date is mainly limited in the green–yellow region. However, the multichip approach is limited by technical and cost issues, which adversely affects its use in general lighting. These costs are higher than those of traditional light sources and phosphor-converted WLEDs because of the complexities of multichip WLEDs, including the required assembly, color mixing, and feedback circuitry to maintain color characteristics. In addition to the need for multiple chips (at least three), the different thermal and degradation properties of the individual LEDs used in the multichip WLEDs require costly feedback and a control system that covers color sensors and circuit drivers. Therefore, this paper proposes a new solution in which a single LED chip that emits three peak wavelengths is combined with a monochromatic phosphor to generate warm white light with a high CRI.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A236
Fig. 3. Typical EL spectra of the proposed vertical InGaN/GaN/Si PRLEDs combined with red phosphor (PRLED + red WLED) and the blue LEDs combined with Ce-doped YAG phosphor (YAG:Ce + blue WLED). The LEDs were driven with a current of 350 mA.
Figure 3 shows the typical EL spectrum obtained from the proposed vertical InGaN/GaN/Si PRLEDs combined with red phosphor (PRLED + red WLED). The typical EL spectrum displayed four peaks at approximately 405, 468, 537, and 620 nm at a driving current of 350 mA. The emission peak at 620 nm mainly originated from the n-UV pumping that excites the red phosphor. Based on this luminescence spectrum, the CCT and average CRI (Ra) were estimated to be 2,900 K and 92, respectively. In this study, eight CRIs (R1– R8) were evaluated to determine the Ra. Conventional WLEDs formed by combining a single blue chip with Ce-doped YAG phosphor (blue + YAG:Ce WLED) were also characterized. The typical EL spectrum exhibited a typical dual-peak luminescence that includes a broad phosphor-converted yellow-green peak and a relatively narrower blue peak from the blue LED, as shown in Fig. 3. The CCT and Ra were estimated to be 5500 K and 70, respectively. The CCT of the PRLED + red WLEDs was lower than that of the blue + YAG:Ce WLEDs, which can be attributed to the fact that the tetrachromatic PRLED + red WLEDs exhibited relatively strong red emission. Although blue + YAG:Ce WLEDs possess inherently high luminous efficiency, they display high CCT and low Ra because they lack a significant red component in the luminescence spectrum. The most convenient approach to reduce and tune the CCT of the blue + YAG:Ce WLEDs is to blend the Ce-doped YAG phosphor with red-emitting phosphor. The CCT and luminous efficiency of the blue + YAG:Ce WLEDs generally decrease when the percentage of red-emitting phosphor incorporated into the YAG:Ce phosphor increases. When the YAG:Ce phosphor is blended with (SrCa)AlSiN3:Eu phosphor and combined with a blue LED at a driving current of 350 mA, the CCT decreases significantly from around 5500 to 2900 K. The luminous efficiency is reduced from 120 to 100 lm/W. The phosphor-converted WLEDs formed by the combination of blue LED chips and YAG:Ce with a red-emitting phosphor exhibited a red-enhanced spectrum, as shown in Fig. 3. The CCT and Ra were estimated to be approximately 2930 K and 73, respectively, in this spectrum. Therefore, CCT can be readily tuned by altering the composition of the phosphor used in the phosphor-converted dichromatic WLEDs. However, increasing CRI to beyond 80 is difficult as a result of the two-color system of the luminescence spectrum of the blue + YAG:Ce WLED. The proposed PRLED + red WLED may be suitable for applications that require high CRI, such as residential and medical lighting, given its low CCT and high CRI, although it displays relatively low luminous efficiency at 50 lm/W [11–13]. The low luminous
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A237
efficiency of the PRLED + red WLEDs can be ascribed to the fact that the EQE of the GaNbased n-UV LEDs used in this study is lower than that of the GaN-based blue LEDs. The InGaN/GaN/Si PRLEDs were based on a non-optimized n-UV LED structure with a typical output power of 330 mW at a driving current of 350 mA. This output power is significantly lower than that of the blue LEDs used in this study, which had a typical output power of 460 mW at a driving current of 350 mA. Aside from optimizing an n-UV LED structure, the blue and green band emission intensities can be enhanced by increasing the absorption α (the fraction of UV light absorbed by the MQW converters) in the blue and green MQWs of PRLEDs. This enhancement boosts the luminous efficiency of the PRLED + red WLEDs. The most straightforward approach for increasing α is to stack additional InGaN-based blue and green QWs in the PRLED. However, this might cause material degradation of the subsequent InGaN/GaN p-i-n heterojunction for n-UV light emission if the total thickness of InGaN layers in the blue and green QWs exceeds the critical thickness. Figure 4 shows the EL spectra obtained from the PRLED + red WLEDs driven by different forward currents. The spectra suggest that the red peak intensity increased more significantly than that of the blue and green peaks with the increase in injection current. This phenomenon can be attributed to the significant drop in the efficiency of the InGaN/GaNbased blue and green QWs as pumping intensity increases with the current injected into the nUV emitting layers, namely, the In0.05Ga0.95N/Al0.1Ga0.9N QWs. This efficiency drop mainly originates from a cubic dependence on charge carrier density under electrical and/or optical excitation, that is, the Auger recombination in the active layer of GaN-based LEDs, which governs the decrease with increasing pumping intensity [2]. This droop effect was slight in short wavelength (n-UV) InGaN LEDs and pronounced at long wavelengths [14]. The red emission peak in the EL spectra of the PRLED + red WLEDs mainly originates from the pumping of n-UV light into the red phosphor. Therefore, the CCT of the PRLED + red WLED decreases with an increase in the currents injected. This decrease of CCT is ascribed to the decrease in the intensity ratio of the blue and green emissions to the red emission. In other words, the efficiencies of the blue and green emissions drop significantly compared with the red emission.
Fig. 4. Typical EL spectra of the PRLED + red WLEDs driven by different forward currents. The inset shows a typical photograph of the PRLED + red WLEDs driven under a current of 100 mA.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A238
Figure 5 depicts the variation of CCT and Ra of typical PRLED + red and BLED + YAG + red WLEDs driven by the different currents. For the PRLED + red WLEDs, the CCT (Ra) was approximately 2920 (92), 2840 (95), and 2770 (93) relative to the injection currents of 350, 700, and 1050 mA, respectively. Although the CCT changed with the injection current in the PRLED + red WLEDs, this variation is small enough to meet the specifications of practical applications, and Ra remains greater than 90.
Fig. 5. Typical CIE color coordinates of the PRLED + red WLEDs driven with different currents.
4. Conclusions This paper presents GaN/Si-based vertical PRLEDs that emit n-UV, blue, and green peaks that are combined with red phosphor to generate white light with a low CCT and CRI. Trichromatic LED chips were combined with red phosphor to generate white light with a CCT and CRI of approximately 2,900 and 92, respectively. Moreover, the CCT of the PRLED + red WLED decreased slightly with the increase in the driving currents injected. This phenomenon can be attributed to the fact that the intensity ratio of the blue and green emissions to the red emission decreased with increased current injection. In other words, the red emission peak that was mainly produced by the pumping of UV light into the red phosphor was slightly influenced by the drop in efficiency of UV LED, unlike the blue and green QWs. Thus, the monolithic PRLED + red WLEDs may be applied to indoor lighting in the future, given the low CCT and CRI of the system. Acknowledgments This work received financial support from the Ministry of Science and Technology Taiwan for financial support under contract Nos. NSC-101-2221-E-218-012-MY3, NSC-101-2221-E006-171-MY3, NSC-100-2112-M-006-011-MY3, and NSC-100-3113-E-006-015.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A239
Abstract: We present a trichromatic GaN-based light-emitting diode (LED) that emits near-ultraviolet (n-UV) blue and green peaks combined with red phosphor to generate white light with a low correlated color temperature (CCT) and high color rendering index (CRI). The LED structure, blue and green unipolar InGaN/GaN multiple quantum wells (MQWs) stacked with a top p-i-n structure containing an InGaN/GaN MQW emitting n-UV light, was grown epitaxially on a single substrate. The trichromatic LED chips feature a vertical conduction structure on a silicon substrate fabricated through wafer bonding and laser lift-off techniques. The blue and green InGaN/GaN MQWs were pumped with n-UV light to re-emit low-energy photons when the LEDs were electrically driven with a forward current. The emission spectrum included three peaks at approximately 405, 468, and 537 nm. Furthermore, the trichromatic LED chips were combined with red phosphor to generate white light with a CCT and CRI of approximately 2900 and 92, respectively. ©2015 Optical Society of America OCIS codes: (230.3670) Light-emitting diodes; (230.0250) Optoelectronics; (230.5590) Quantum-well, -wire and -dot devices.
References and links 1. 2. 3. 4. 5. 6.
E. F. Schubert and J. K. Kim, “Solid-state light sources getting smart,” Science 308(5726), 1274–1278 (2005). G. Verzellesi, D. Saguatti, M. Meneghini, F. Bertazzi, M. Goano, G. Meneghesso, and E. Zanoni, “Efficiency droop in InGaN/GaN blue light-emitting diodes: physical mechanisms and remedies,” J. Appl. Phys. 114(7), 071101 (2013). J. H. Oh, J. R. Oh, H. K. Park, Y. G. Sung, and Y. R. Do, “New paradigm of multi-chip white LEDs: combination of an InGaN blue LED and full down-converted phosphor-converted LEDs,” Opt. Express 19(S3 Suppl 3), A270–A279 (2011). K. J. Chen, H. T. Kuo, H. C. Chen, M. H. Shih, C. H. Wang, S. H. Chien, S. H. Chiu, C. C. Lin, C. J. Pan, and H. C. Kuo, “High thermal stability of correlated color temperature using current compensation in hybrid warm white high-voltage LEDs,” Opt. Express 21(S2 Suppl 2), A201–A207 (2013). C. C. Lin and R. S. Liu, “Advances in phosphors for light-emitting diodes,” J. Phys. Chem. Lett. 2(11), 1268– 1277 (2011). J. Y. Tsao, M. H. Crawford, M. E. Coltrin, A. J. Fischer, D. D. Koleske, G. S. Subramania, G. T. Wang, J. J. Wierer, and R. F. Karlicek, Jr., “Toward smart and ultra-efficient solid-state lighting,” Adv. Opt. Mater. 2(9), 1– 28 (2014).
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A232
7. 8. 9. 10. 11. 12. 13. 14.
J. M. Phillips, M. E. Coltrin, M. H. Crawford, A. J. Fischer, M. R. Krames, R. Mueller-Mach, G. O. Mueller, Y. Ohno, L. E. S. Rohwer, J. A. Simmons, and J. Y. Tsao, “Research challenges to ultra-efficient inorganic solidstate lighting,” Laser Photon. Rev. 1(4), 307–333 (2007). J. K. Sheu, F. B. Chen, S. H. Wu, M. L. Lee, P. C. Chen, and Y. H. Yeh, “Vertical InGaN-based green-band solar cells operating under high solar concentration up to 300 suns,” Opt. Express 22(S5 Suppl 5), A1222–A1228 (2014). J. K. Sheu, S. J. Chang, C. H. Kuo, Y. K. Su, L. W. Wu, Y. C. Lin, W. C. Lai, J. M. Tsai, G. C. Chi, and R. K. Wu, “White-light emission from near UV InGaN-GaN LED chip precoated with blue/green/red phosphors,” IEEE Photon. Technol. Lett. 15(1), 18–20 (2003). D. Schiavon, M. Binder, A. Loeffler, and M. Peter, “Optically pumped GaInN/GaN multiple quantum wells for the realization of efficient green light-emitting devices,” Appl. Phys. Lett. 102(11), 113509 (2013) (and references therein). Y. Chen, M. Gong, G. Wang, and Q. Su, “High efficient and low color-temperature white light-emitting diodes with Tb3 Al5O12: Ce3+ phosphor,” Appl. Phys. Lett. 91(7), 071117 (2007). H. S. Jang, W. B. Im, D. C. Lee, D. Y. Jeon, and S. S. Kim, “Enhancement of red spectral emission intensity of Y3Al5O12:Ce3+ phosphor via Pr co-doping and Tb substitution for the application to white LEDs,” J. Lumin. 126(2), 371–377 (2007). S. Nizamoglu, T. Erdem, X. W. Sun, and H. V. Demir, “Warm-white light-emitting diodes integrated with colloidal quantum dots for high luminous efficacy and color rendering,” Opt. Lett. 35(20), 3372–3374 (2010). J. R. Chen, Y. C. Wu, S. C. Ling, T. S. Ko, T. C. Lu, H. C. Kuo, Y. K. Kuo, and S.-C. Wang, “Investigation of wavelength-dependent efficiency droop in InGaN light-emitting diodes,” Appl. Phys. B 98(4), 779–789 (2010).
1. Introduction White light-emitting diodes (WLEDs) have been recognized as a promising next-generation general lighting source in response to the issues of energy shortages and environmental degradation [1]. Commercial WLEDs are produced mainly in two ways. The first selectively combines red, green, and blue (RGB) LEDs. However, this is costly, and the overall light outputs of RGB LEDs degrade at different rates. As a result, the color is eventually unbalanced. The most popular method of producing WLEDs is by combining GaN-based blue LEDs and yellow phosphor. Another possible technique of white light generation involves the combination of deep ultraviolet (UV) LEDs and RGB phosphors that act as fluorescent tubes. Nonetheless, the development of deep UV-pumped WLEDs remains in its infancy because of the limited efficiency of deep UV LEDs. The efficiency of GaN-based LEDs depends strongly on the indium content in the active layer. InGaN LEDs exhibit high external quantum efficiencies (EQEs) in the violet and blue regions. However, this efficiency is significantly lower in green emission devices, which have greater indium content in the InGaN active region. This efficiency loss can be attributed to a number of material issues, including highdensity dislocation and strong polarization fields [2]. Commercial WLEDs are mainly produced through the combination of a single blue (λ = 440–460 nm) LED chip with yellow phosphor (λ ~560 nm), such as a cerium-doped yttrium aluminum garnet (YAG:Ce), in consideration of cost-performance merits. Their efficiency has far exceeded that of traditional incandescent bulbs. Dichromatic WLEDs with a color rendering index (CRI) of approximately 70–80 and a correlated color temperature (CCT) of 4000–8000 K are suitable for outdoor applications [1]; however, these LEDs are inadequate for indoor illumination applications that generally require CRIs higher than 80 [3]. To improve color characteristics, a second red (or orange) LED chip or a red-emitting phosphor has been integrated to generate a trichromatic WLED [4,5]. Similarly, the CRI can be boosted to a maximum of 85 through the multichip approach, which uses LED chips alone to generate white light with color components at red, green, and blue wavelengths [6]. Another multichip approach utilizes four LED chips that emit red, yellow, green, and blue (RYGB) wavelengths to establish a tetrachromatic WLED that can achieve a CRI higher than 95 [7]. This study introduces a trichromatic GaN-based LED that emits near-UV (n-UV), blue, and green peaks combined with red phosphor to generate white light with a low CCT and high CRI. In the spectrum of this trichromatic LED, the blue and green peaks originate from photon recycling using n-UV light excitation, which is generated by electrical pumping, to optically pump two separate multiple quantum wells (MQWs). Blue and green unipolar InGaN/GaN MQW structures are
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A233
grown on the same substrate, which is stacked with a top p-i-n structure containing an n-UVemitting InGaN/GaN MQW. The blue and green InGaN/GaN MQWs are pumped with n-UV light to re-emit low-energy photons when the n-UV MQW is electrically driven by a forward current. The resulting emission spectrum shows three peaks at approximately 405, 468, and 537 nm. The trichromatic LED chips are then combined with red phosphor to generate white light. The trichromatic GaN-based LEDs have a vertical conduction structure on a silicon substrate that is fabricated using wafer bonding and laser lift-off (LLO) techniques to alleviate the thermal effect on performance degradation when the devices operate at high input power [8]. The fabrication procedures and characterizations of the experimental LEDs are detailed in the following sections. 2. Experiments The InGaN-based LEDs used in this study were grown on c-face sapphire substrates using a metalorganic vapor-phase epitaxy reactor. The layer structure consisted of a 30-nm-thick lowtemperature GaN nucleation layer and 3-µm-thick undoped GaN (u-GaN) with a chamber pressure of 500 Torr. A 16-pair green InGaN/GaN MQW containing a 2.5-nm-thick In0.28Ga0.72N well separated by a 9-nm-thick GaN barrier in each pair was also grown in sequence with a 16-pair blue InGaN/GaN MQW containing a 2-nm-thick In0.15Ga0.85N well separated by a 8-nm-thick GaN barrier. The green and blue MQWs emitted light wavelengths of approximately 535 nm and 465 nm, respectively, as determined by photoluminescence (PL) measurements. A p–i–n heterostructure was then grown sequentially on the green and blue MQWs. The heterostructure consisted of a 4-µm-thick Si-doped n+-GaN layer (n ~3 × 1018 cm−3), a 12-pair MQW that contained twelve 3-nm-thick In0.05Ga0.95N wells separated by 12-nm-thick Al0.15Ga0.85N barriers, and a 200-nm-thick Mg-doped p-GaN layer (p ~5 × 1017 cm−3). The typical PL peak wavelength, as obtained from the In0.05Ga0.95N/Al0.15Ga0.85N MQW, was approximately 405 nm, and this MQW was denoted as an n-UV violet MQW. Figure 1(a) shows the schematic of the layer structure of the aforementioned epitaxial heterostructure. A bilayer Ni/Ag metal (1/200 nm) was deposited onto the top p-GaN layer following epitaxial growth to serve as a reflector/ohmic contact layer [8]. Once the reflector layer was formed, a barrier layer consisting of a 200-nm-thick TiW layer and a 50-nm-thick Pt layer was configured to alleviate the diffusion of Ag from the reflector layer to the bonding layer. This barrier layer was deposited between the reflector and the bonding layers. A 3-µmthick In layer was used as a bonding layer in the proposed vertical InGaN/GaN/Si photon recycling LEDs (PRLEDs). In addition, the Si substrates were dipped into a buffered oxide etch solution for 60 s to remove the native silicon dioxide; a bilayer Ti/Au (20/1500 nm) metal was then deposited onto the surfaces of the Si substrates to serve as ohmic contact and bonding layers [8]. These Si wafers acted as receptors during the wafer bonding process. The sapphire substrate was removed by LLO to expose the u-GaN layer after the wafer bonding process. The InxGa1−xN/GaN-based heteroepitaxial layers were thus transferred to the Si substrate along with the u-GaN top layer. The samples were then treated in a potassium hydroxide solution (concentration of 3 M) at an elevated temperature of 60°C to texture the uGaN layer for enhancing light extraction efficiency. The n+-GaN layer was exposed by selective-area dry etching after surface texturing. Ti/Al/Ni/Au (20/30/150/2000 nm) metal layers were then deposited onto this exposed n+-GaN layer to form n-type ohmic contacts (cathode electrodes) on the wafers. Finally, the Si substrates were thinned to 150 µm and coated with Ti/Au (50/500 nm) metals to serve as the rear ohmic contact layer.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A234
Fig. 1. (a) Layer structure of the PRLED epitaxial heterostructure. (b) Typical top-view image of the proposed vertical InGaN/GaN/Si PRLED chip. (c) Schematic cross-section layer structure of the vertical PRLED chip.
Figure 1(b) shows a typical top-view image of the proposed vertical InGaN/GaN/Si PRLEDs as obtained through an optical microscope. The dashed line indicates the cross section that corresponds to a schematic of the layer structure of the vertical LEDs, as shown in Fig. 1(c). All of the experimental PRLED chips used in the present study had an area of 1.25 × 1.25 mm2. To generate the WLEDs, the trichromatic PRLED chips were coated with red phosphor that was mixed with resin. The red phosphor was a Eu-doped (SrCa)AlSiN3 powder with an average particle size of 10 μm. The room-temperature current-voltage (I—V) characteristics of the experimental LEDs were measured using a HP-4156C semiconductor parameter analyzer. The light output power-current (L—I) of the LEDs was measured using a calibrated integrating sphere that was combined with a Keithly 2400 source meter. 3. Results and discussions Figure 2(a) depicts the typical I—V characteristics of the vertical InGaN/GaN/Si PRLEDs and the n-UV LEDs without the stacked optical converters. The forward I—V curves of the LEDs do not differ significantly. In principle, a fraction of pump light, that is, the n-UV light emission generated by direct current injection, is absorbed by the optical converters (i.e., the stacked blue and green InGaN quantum wells). This fraction increases with the total thickness of the converters. However, the lattice mismatch between GaN and InxGa1−xN limits the total thickness of the InxGa1−xN layer in the optical converters of the PRLED structure, particularly when the In content is increased. In other words, the performance of the p–i–n GaN/InxGa1−xN electrical n-UV emitter declines if the total thickness and/or In content of the optical converters are too high, because the n-UV emitter is grown on the optical converters, as indicated in Fig. 1(a). Considering the I—V characteristics, Fig. 2(a) suggests that the epitaxial growth of the p–i–n heterostructure on the optical converters (blue and green MQWs) did not significantly degrade in material quality.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A235
Fig. 2. (a) Typical I-V characteristics of vertical InGaN/GaN/Si PRLEDs and n-UV LEDs without the stacked optical converters. (b) Typical EL spectra of the vertical InGaN/GaN/Si PRLEDs and n-UV LEDs without the stacked optical converters, under an injection current of 350 mA.
Figure 2(b) displays the typical electroluminescence (EL) spectra obtained from vertical InGaN/GaN/Si PRLEDs and the n-UV LEDs without stacked optical converters under an injection current of 350 mA. As expected, the n-UV MQW exhibited a peak at 405 nm because of carrier injection into the In0.05Ga0.95N/Al0.1Ga0.9N MQW and the recombination therein. Aside from this n-UV peak, the PRLEDs displayed two additional peaks at 468 nm and 537 nm, which originated from the optical pumping with current-injected n-UV light. This pumping excited the stacked blue and green MQWs. Nonetheless, the phosphorconverted emissions exhibited broader spectra than the light emissions of the direct currentinjected LEDs or the photon-recycling semiconductors [10], although n-UV LEDs have been shown to excite phosphors through down conversion to generate distinct RGB emission peaks [9]. Luminous efficacy is limited in traditional white light sources based on emissions with broad spectra, because the sensitivity of the human eye declines rapidly as one approaches infrared and UV wavelengths. In addition, the human eye possesses only three types of colorsensitive receptors or cones. Therefore, high-efficiency WLEDs can be produced by generating light with three distinct colors as with the multichip approach, which produces white light using RGB LEDs [1]. Moreover, the assembly of InGaN-based blue and green LEDs and of the AlGaInP-based red LED can enhance luminous efficiency more than other solutions can, although LED performance to date is mainly limited in the green–yellow region. However, the multichip approach is limited by technical and cost issues, which adversely affects its use in general lighting. These costs are higher than those of traditional light sources and phosphor-converted WLEDs because of the complexities of multichip WLEDs, including the required assembly, color mixing, and feedback circuitry to maintain color characteristics. In addition to the need for multiple chips (at least three), the different thermal and degradation properties of the individual LEDs used in the multichip WLEDs require costly feedback and a control system that covers color sensors and circuit drivers. Therefore, this paper proposes a new solution in which a single LED chip that emits three peak wavelengths is combined with a monochromatic phosphor to generate warm white light with a high CRI.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A236
Fig. 3. Typical EL spectra of the proposed vertical InGaN/GaN/Si PRLEDs combined with red phosphor (PRLED + red WLED) and the blue LEDs combined with Ce-doped YAG phosphor (YAG:Ce + blue WLED). The LEDs were driven with a current of 350 mA.
Figure 3 shows the typical EL spectrum obtained from the proposed vertical InGaN/GaN/Si PRLEDs combined with red phosphor (PRLED + red WLED). The typical EL spectrum displayed four peaks at approximately 405, 468, 537, and 620 nm at a driving current of 350 mA. The emission peak at 620 nm mainly originated from the n-UV pumping that excites the red phosphor. Based on this luminescence spectrum, the CCT and average CRI (Ra) were estimated to be 2,900 K and 92, respectively. In this study, eight CRIs (R1– R8) were evaluated to determine the Ra. Conventional WLEDs formed by combining a single blue chip with Ce-doped YAG phosphor (blue + YAG:Ce WLED) were also characterized. The typical EL spectrum exhibited a typical dual-peak luminescence that includes a broad phosphor-converted yellow-green peak and a relatively narrower blue peak from the blue LED, as shown in Fig. 3. The CCT and Ra were estimated to be 5500 K and 70, respectively. The CCT of the PRLED + red WLEDs was lower than that of the blue + YAG:Ce WLEDs, which can be attributed to the fact that the tetrachromatic PRLED + red WLEDs exhibited relatively strong red emission. Although blue + YAG:Ce WLEDs possess inherently high luminous efficiency, they display high CCT and low Ra because they lack a significant red component in the luminescence spectrum. The most convenient approach to reduce and tune the CCT of the blue + YAG:Ce WLEDs is to blend the Ce-doped YAG phosphor with red-emitting phosphor. The CCT and luminous efficiency of the blue + YAG:Ce WLEDs generally decrease when the percentage of red-emitting phosphor incorporated into the YAG:Ce phosphor increases. When the YAG:Ce phosphor is blended with (SrCa)AlSiN3:Eu phosphor and combined with a blue LED at a driving current of 350 mA, the CCT decreases significantly from around 5500 to 2900 K. The luminous efficiency is reduced from 120 to 100 lm/W. The phosphor-converted WLEDs formed by the combination of blue LED chips and YAG:Ce with a red-emitting phosphor exhibited a red-enhanced spectrum, as shown in Fig. 3. The CCT and Ra were estimated to be approximately 2930 K and 73, respectively, in this spectrum. Therefore, CCT can be readily tuned by altering the composition of the phosphor used in the phosphor-converted dichromatic WLEDs. However, increasing CRI to beyond 80 is difficult as a result of the two-color system of the luminescence spectrum of the blue + YAG:Ce WLED. The proposed PRLED + red WLED may be suitable for applications that require high CRI, such as residential and medical lighting, given its low CCT and high CRI, although it displays relatively low luminous efficiency at 50 lm/W [11–13]. The low luminous
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A237
efficiency of the PRLED + red WLEDs can be ascribed to the fact that the EQE of the GaNbased n-UV LEDs used in this study is lower than that of the GaN-based blue LEDs. The InGaN/GaN/Si PRLEDs were based on a non-optimized n-UV LED structure with a typical output power of 330 mW at a driving current of 350 mA. This output power is significantly lower than that of the blue LEDs used in this study, which had a typical output power of 460 mW at a driving current of 350 mA. Aside from optimizing an n-UV LED structure, the blue and green band emission intensities can be enhanced by increasing the absorption α (the fraction of UV light absorbed by the MQW converters) in the blue and green MQWs of PRLEDs. This enhancement boosts the luminous efficiency of the PRLED + red WLEDs. The most straightforward approach for increasing α is to stack additional InGaN-based blue and green QWs in the PRLED. However, this might cause material degradation of the subsequent InGaN/GaN p-i-n heterojunction for n-UV light emission if the total thickness of InGaN layers in the blue and green QWs exceeds the critical thickness. Figure 4 shows the EL spectra obtained from the PRLED + red WLEDs driven by different forward currents. The spectra suggest that the red peak intensity increased more significantly than that of the blue and green peaks with the increase in injection current. This phenomenon can be attributed to the significant drop in the efficiency of the InGaN/GaNbased blue and green QWs as pumping intensity increases with the current injected into the nUV emitting layers, namely, the In0.05Ga0.95N/Al0.1Ga0.9N QWs. This efficiency drop mainly originates from a cubic dependence on charge carrier density under electrical and/or optical excitation, that is, the Auger recombination in the active layer of GaN-based LEDs, which governs the decrease with increasing pumping intensity [2]. This droop effect was slight in short wavelength (n-UV) InGaN LEDs and pronounced at long wavelengths [14]. The red emission peak in the EL spectra of the PRLED + red WLEDs mainly originates from the pumping of n-UV light into the red phosphor. Therefore, the CCT of the PRLED + red WLED decreases with an increase in the currents injected. This decrease of CCT is ascribed to the decrease in the intensity ratio of the blue and green emissions to the red emission. In other words, the efficiencies of the blue and green emissions drop significantly compared with the red emission.
Fig. 4. Typical EL spectra of the PRLED + red WLEDs driven by different forward currents. The inset shows a typical photograph of the PRLED + red WLEDs driven under a current of 100 mA.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A238
Figure 5 depicts the variation of CCT and Ra of typical PRLED + red and BLED + YAG + red WLEDs driven by the different currents. For the PRLED + red WLEDs, the CCT (Ra) was approximately 2920 (92), 2840 (95), and 2770 (93) relative to the injection currents of 350, 700, and 1050 mA, respectively. Although the CCT changed with the injection current in the PRLED + red WLEDs, this variation is small enough to meet the specifications of practical applications, and Ra remains greater than 90.
Fig. 5. Typical CIE color coordinates of the PRLED + red WLEDs driven with different currents.
4. Conclusions This paper presents GaN/Si-based vertical PRLEDs that emit n-UV, blue, and green peaks that are combined with red phosphor to generate white light with a low CCT and CRI. Trichromatic LED chips were combined with red phosphor to generate white light with a CCT and CRI of approximately 2,900 and 92, respectively. Moreover, the CCT of the PRLED + red WLED decreased slightly with the increase in the driving currents injected. This phenomenon can be attributed to the fact that the intensity ratio of the blue and green emissions to the red emission decreased with increased current injection. In other words, the red emission peak that was mainly produced by the pumping of UV light into the red phosphor was slightly influenced by the drop in efficiency of UV LED, unlike the blue and green QWs. Thus, the monolithic PRLED + red WLEDs may be applied to indoor lighting in the future, given the low CCT and CRI of the system. Acknowledgments This work received financial support from the Ministry of Science and Technology Taiwan for financial support under contract Nos. NSC-101-2221-E-218-012-MY3, NSC-101-2221-E006-171-MY3, NSC-100-2112-M-006-011-MY3, and NSC-100-3113-E-006-015.
#227224 - $15.00 USD (C) 2015 OSA
Received 19 Nov 2014; revised 16 Feb 2015; accepted 18 Feb 2015; published 25 Feb 2015 6 Apr 2015 | Vol. 23, No. 7 | DOI:10.1364/OE.23.00A232 | OPTICS EXPRESS A239
