Open Access
Add to BibSonomy
Abstract
A novel method to stabilize the correlated color temperature in pcW-LEDs from their initial turn-on state to thermal equilibrium is proposed and demonstrated. Under the normal operation condition, it can stabilize the CCT of a pcW-LED by the positive matching of the blue LED peak wavelength to the phosphor excitation spectrum. When the operating temperature unavoidably becomes higher in the LED die quickly after the initial turn-on, the phosphor conversion efficiency degrades and the LED blue light performs red shift. With the positive matching, the red shift actually helps enhance the excitation efficiency of the phosphor to compensate the thermal quenching and efficiency degradation. Therefore, the ratio of the blue light to the yellow light can keep almost constant, as well as the CCTs. In the experiments, the CCT variation could be as small as from 7 K to 83 K in different cases. Finally, we introduce a new factor, the so-called guide number, which is used to count the total change of the enhancement in equivalent excitation efficiency and the relative reduction of the phosphor light emission. The guide number essentially helps in designing the matching blue LED die and phosphor pair for good CCT stabilization.
© 2017 Optical Society of America
1. Introduction
Phosphor-converted white LEDs (pcW-LEDs) have been the best way in solid-state lighting to generate white light for its low cost and high reliability. pcW-LED is mainly based on a LED die, which emits the blue light, covered with one-phosphor or two-phosphor layer to convert the partly blue light into the yellow light or the green and red lights, respectively [1–4]. A proper recipe of the phosphors makes the white light emitting at a specific correlated color temperature (CCT) with the chromaticity coordinates close to the Planckian locus. Although two-phosphor recipe performs better in color rendition, using one-phosphor (yellow) is the simplest way with the lowest cost but still acceptable color-rendering index (CRI). This is why the pcW-LED with yellow phosphor has dominated the solid-state light sources.
When evaluating a light source, we consider not only the characteristics in electricity, but also including the features in photometry and chromaticity [5]. The latter two are more critical than the others are, because these two critical characteristics are very sensitive to the users’ eyes. As the matter of fact, the low luminous flux cannot meet with the general illumination requirement, and the improper chromatic performance effects both in psychology and in physiology [6]. Indeed, the pcW-LED is a complicated system rather than a common electro-optical device. Moreover, the external quantum efficiency of the blue LED die and the quantum efficiency of the phosphor are not perfect. In the energy viewpoint, it means that always there is conversion from part of the injection power into the heat whenever it is in operation. The generated heat increases the temperature in both blue die and phosphor. In the most practical design, no matter where the phosphor location is, including conformal coating and remote structure [7], the phosphor temperature is always comparable with that in the blue die [8-9]. As a result, the thermal effect in the pcW-LED system plays an important role of an uncertain factor, which could affect the performance of pcW-LEDs.
As a GaN-based die operating at high current density, there always exists competition between the blue and the red shift mechanisms [10–12]. The blue shift results from the band filling effect and the carrier screening of the quantum-confined Stark effect (QCSE) [13]. Nonpolar or semi-polar GaN is helpful to alleviate QCSE [14], but the approach requires sophisticated growth techniques or freestanding GaN substrates [15]. Either of them is not commercially favorable because of the high production cost. Meanwhile, the red shift originates from the thermally induced bandgap reduction [16], which is inherent in semiconductor materials. Such effect is an inevitable nature of semiconductor materials. Usually a pcW-LED operates under the constant injection current, and the thermal effect will cause obvious red shift from the initial state to thermal equilibrium.
In spite of low impact on the absorption band of the phosphor, the thermal effect does bring high impact on the conversion efficiency of the phosphor. Once the phosphor’s temperature increases, the thermal quenching of the phosphor particles reduces the external quantum efficiency such that the emission of the yellow light is lesser and the CCT of pcW-LED then goes high. The United States Department of Energy (DOE) has discovered that the blue light from a general pcW-LED is at the similar level to the conventional light sources, and does not causes additional optical safety issues to users [17]. A pcW-LED, however, needs to keep good balance between the blue light and yellow light under the normal operation as maintaining the setting CCT. Otherwise, it would induce heavy CCT drift or even blue light leakage in the worst condition. In this paper, we discover a novel passive scheme for a well match design between the blue die and the phosphor for the unavoidable thermal effect. The novel approach makes the self-compensation between the blue and yellow lights under the normal operation, and then the CCT keeps stabilized.
2. Method
Upon thermal quenching, the external quantum efficiency (EQE) of phosphor drops such as to causes CCT of pcW-LED to drift. To stabilize the CCT of a pcW-LED, there are two mechanisms taken into consideration. One is the reduction of the phosphor conversion efficiency due to the thermal effect. The other is the enhancement of phosphor excitation efficiency due to the red shift of the LED blue light from LED by the thermal effect. As well balancing between these two mechanisms, the self-compensation process will stabilize the CCT of pcW-LED automatically. Figure 1 shows the excitation spectrum of the phosphor and the LED pumping lights before (30°C) and after (120°C) the red shift. It illustrates how the conversion efficiency of a phosphor depends on the pumping wavelength.
Fig. 1 Red shift spectrum of the blue die vs. the excitation spectrum of the phosphor. The pumping wavelength is located at the spectrum of (a) positive slope, and (b) negative slope. (c) The pumping spectra of the blue dies in the following experiment.
Generally, the emission wavelength of the blue die fits to the maximum of the excitation spectrum of the phosphors to receive the maximal conversion efficiency. If so, the LED die is of negative matching to the phosphor [Fig. 1(b)], which the red shift causes the decrease of the excitation efficiency and leads to CCT drifting. In contrast, as the LED die is of positive matching to the phosphor [Fig. 1(a)], the red shift of the LED die then cause increase of the excitation efficiency that compensates the reduction of the phosphor emission by the thermal effect. The proposed scheme is to choose the blue dies with LED peak wavelengths located at the positive slope regime of the phosphor excitation spectrum, the LED red shift spectrum upon thermal effect actually results in higher phosphor excitation efficiency, as shown in Fig. 1(a). In such a way, the higher phosphor excitation efficiency compensates to the drop of quantum efficiency of the phosphor by thermal quenching. Unlike the very difficult situation of the color binning for the applications in display backlights, there already find several kinds of high efficient phosphors with positive matching to the common blue light LED dies. For examples, YAG with maximum of absorption ranging from 440 nm to 460 nm, CaS-hosted phosphor (460 nm - 470 nm) [18], Sr3SiO5:Ce3+ (410 nm - 430 nm) [19], Sr3AlSiO5:Eu2+ (410 nm - 420 nm) [20], Mg3Gd2Ge3O12: Ce3+ (460 nm - 470 nm), Ca(Si, Al)N2: Ce3+ (460 nm - 470 nm), and nitride phosphor (440 nm - 450 nm) [21]. Taking YAG with a maximum of excitation efficiency at 460 nm as an example, we choose several blue LED dies with the emission peak wavelength from 422 nm to 459 nm for the following experiment.
To figure out the change of the excitation efficiency of the blue lights upon red shift, an equivalent excitation efficiency () expresses as
where is the normalized blue light spectrum emitted from the blue die, and is the normalized excitation spectrum of the phosphor. The red shift of the blue die should cause increasing of for the positive matching cases, vice versa for the negative matching ones.In Fig. 2, all the 13 cases in experiments present the various CCT drifting of the pcW-LEDs. Some cases are of CCT around 4600K (neutral white), and the others are of CCT around 6500K (cool white). These pcW-LEDs are using the hemispherical phosphor packaging design with 6.8% and 10% of phosphor weight concentration, respectively. As driving pcW-LEDs at dc 350 mA, we recorded the backside board temperature and the forward voltage of pcW-LED extensively. All the photometric characteristics were studied as the backside board temperature increased from 30°C to 120°C. Obviously, the measured thermal-chromatic characteristics of those pcW-LEDs can categorize into three groups. The first group is for the pcW-LEDs with blue peak wavelengths around 446 nm or longer. The feature is their CCTs increase as the pcW-LED heating up. The second group is for the pcW-LEDs with blue peak wavelengths around 437 nm or shorter. The feature is their CCTs decrease as the pcW-LED heating up. The third group, also a self-compensation group, is for the pcW-LEDs with blue peak wavelengths between 437 nm and 444 nm. The feature is their CCTs always decreases first and then turns to increases as the pcW-LEDs heating up. When these samples reach the thermal equilibria, the final CCTs will not drift far away from their initial values.
Fig. 2 The chromaticity drifting curves and the CCT drifts of all the 13 experimental cases. Some cases are of CCT around 4600K (neutral white), and the others are of CCT around 6500K (cool white). The initial and final temperature are 30° and 120°, respectively. 6.8% and 10% indicate the weight concentration of the phosphors applied to the samples.
Here is the story inside. In the first group, the LED dies are negative matching to the YAG phosphors, which the blue LED peak wavelengths are located at peak or the negative slope regime of the phosphor excitation spectrum. The thermal-induced red shift of the blue LED die will cause decrease of the excitation efficiency. Then, the pcW-LED performs of lesser yellow light emission and higher CCT as the temperature increasing. On the other hand, the blue LED peak wavelengths of the second group are located at the steep positive slope regime and far from the maximum of the phosphor excitation spectrum. In such a way, the thermal-induced red shift of the blue die will cause increase of the excitation efficiency greater than the reduction by the phosphor thermal quenching of the phosphor. Then, the pcW-LED performs of more yellow light emission and lower CCT as the temperature increasing. As to the third group, the blue LED peak wavelength is located at the flatter positive slope regime of the phosphor excitation spectrum. When turning on the tested pcW-LEDs, the thermal-induced red shift happens in the similar way again. In the beginning, the phosphor excitation efficiency increases. Then, like the cases in the second group, the pcW-LED performs of more yellow light emission and lower CCT as the temperature increasing. The situation changes after the thermal-induced red shift comes to the negative matching like the cases in the first group. Then, the CCT turns to increase. Based on the feature of the third group, the thermal-induced CCT drift can be well controlled in an acceptable range as long as proper selection of the LED die and the phosphor. In the experiment, an LED die with peak wavelength of 443 nm packaging with 10% (6.8%) weight concentration of phosphor performs its CCT drift as small as 7K (83K), while its initial CCT around 4700K (6300K). As shown in Fig. 3, to stabilize CCT by self-compensation is feasible for the normal operation.
Fig. 3 CCT drift according to temperature increase. (a) The initial CCT is 4700K and phosphor concentration is 10%; (b) the initial CCT is 6300K and the phosphor concentration is 6.8%. Δxy is the distance in the color coordinate from the point at 30°C to 120°C.
3. Quantitative analysis
With aid of Eq. (1), we calculate the change of , and compare the thermal degradation of the phosphor for further studying the CCT variation behavior quantitatively. The measurement on the thermal degradation of the phosphor utilizes a blue laser to illuminate on a 1.5mm-thickness phosphor plate that attaches on a temperature-controlled substrate. An IR camera monitored the surface temperature of the phosphor plate. As shown in Fig. 4, the concentration of the phosphor was 6.8% (or 10%) to have the yellow light decay of 0.65% (or 1.86%) relative to its initial value from 30°C to 120°C. The pcW-LEDs in the experiment were prepared by packaging the blue LEDs dies covered with the dome-shaped phosphors of 1.5 mm in radius and a hemispherical lens on the phosphor dome. In the meantime, a thermal coupler sensor also monitors the backside board temperature of the pcW-LED. Actually, the real phosphor temperature in the pcW-LED is difficult to measure, and the temperature distribution is not uniform across over whole the phosphor volume. It is practical, however, to compare the change of the phosphor excitation efficiency due to the red shift of the blue light in term of the backside board temperature, with the thermal degradation of the phosphor as shown in Fig. 4. Figures 5-7 and Tabs. 1-3 show the results of the first group, second group, and the third group, respectively. is the normalized deviation of from 30°C to 120°C for the blue LED die; and is the normalized deviation of the yellow light emission for the phosphor plate from 30°C to 120°C. Finally, the total sum of and defines a guide number (G#) for effectively evaluating the degree of the self-compensation for the CCT stabilization in pcW-LEDs. More close to zero the G# is, the better the self-compensation is.
Fig. 5 CCT monitoring from 30°C to 120°C on the board for the first group. The figure shows that the CCT moves lower when the temperature rises.
Fig. 6 CCT monitoring from 30°C to 120°C on the board for the second group The figure shows that the CCT moves higher when the temperature rises.
Fig. 7 CCT monitoring from 30°C to 120°C on the board for the third group. A turning point emerges in the figure for each pcW-LED.
Table 1. The parameters and guide numbers corresponding to Fig. 5. Calculation of the guide number as the board temperature raising from 30°C to 120°C for the second group.
Table 2. The parameters and guide numbers corresponding to Fig. 6. Calculation of the guide number as the backside board temperature raising from 30°C to 120°C for the first group.
Table 3. The parameters and guide numbers corresponding to Fig. 7. Calculation of the guide number as the backside board temperature raising from 30°C to 120°C for the first group.
Figure 5 shows the results for the first group, where the peak wavelength of the blue light is shorter than other groups. Since the thermal-induced red shift enables big gain in rather than the smaller loss in, the guide number G# is positive and of large value. The corresponding CCT change is moving along to lower CCT due to more yellow light emission. In contrast, Fig. 6 shows the result for the second group, where the peak wavelength of the blue light is longer than other groups. The thermal-induced red shift causes more loss in than the gain from increasing, so the G# is a large negative value. The corresponding CCT moves along to higher CCT. Such a condition could trigger more concerns about the blue light leakage issue. Figure 7 is for the third group, which is easily to observe the self-balancing behavior. Except the pcW-LED of wavelength of 437 nm with 6.8% of phosphor concentration, the other three pcW-LEDs have the guide numbers near zero, and a turning point emerges in each pcW-LED.
Figure 8 depicts the relation between CCT and G#. The phosphor concentration becomes to two segments, i.e., higher and lower phosphor concentrations. Higher concentration (10%) is for the segment of lower CCT, which were between 4600K and 5100K in the experiment, and are marked as yellow spots in Fig. 8. Phosphor concentration of 6.8% is for the segment of higher CCT, which were between 6400K and 8000K. Large G#, with both positive and negative, perform obvious CCT drift. These cases include CCT of 2204K for peak wavelength of 422 nm in the first segment, CCT of 1230K (1144K) for the peak wavelength of 456 nm (459 nm) in the second segment. The cases of G# close to zero perform lesser CCT drift, and some of them behave with the existence of the turning points.
Fig. 8 CCT drift from 30°C to 120°C vs. the guide number. The yellow (blue) circles/triangles are for phosphor concentration of 10% (6.8%). The triangles are for the cases with observation of turning points during CCT drift.
To simulate the aging behaviors of pc-WEDs by thermal effect, the experiments executed accelerating aging tests by increasing the injection current. Figure 9 shows the output flux and CCT variation as a function of aging hour in the case of the CCT around 6500 K. The measurement consists of two stages. The first stage was to measure the change in flux and CCT when the pcW-LED reached thermal equilibrium from the turn-on moment while the injection current were kept 0.8 A. The cases for peak wavelength of 441 nm and 444 nm performed less CCT drift compared to others when reaching thermal equilibrium. Obviously, as indicated in our approach, proper selection of the phosphor and the peak wavelength of the blue LED can keep blue-yellow balance in the normal operation period of a pcW-LED. In the second stage, we observed the CCT drift when the optical flux drops below 70% of that in thermal equilibrium. We found that the self-balance mechanism was not able to prevent the dramatic increase of CCT when the pcW-LED is in the break down process, when the whole pcW-LED operated in the too high temperature condition owing to serious aging happening. In this stage, it suggests the new pcW-LED replacement.
Fig. 9 The variation of both the luminous flux (green thing line) and CCTs (purple thin line) during the aging test, which is corresponding to their spectra as comparing to the excitation spectrum of the phosphor. The blue circles, green squares, and the red crosses indicate the initial, the equilibrium, and the L70 states, respectively.
4. Conclusion
In this paper, we have proposed and demonstrated a novel method for stabilizing CCTs by self-compensation of phosphor efficiency in pcW-LEDs from turn-on state to thermal equilibrium. The study proposes to select phosphor with a positive slope matching to the excitation spectrum. We suggest locating the peak wavelength of the blue LED die within the positive slope regime of the phosphor excitation spectrum. When the temperature rises with the LED die due to limited external quantum efficiency, the emission light performs red shift, which will cause higher excitation rate of the phosphor to compensate the thermal quenching and efficiency degradation of the phosphor. Therefore, the ratio of the blue to the yellow light keeps almost constant.
Based on the matching of the blue light peak wavelength to the phosphor excitation spectrum, the pcW-LEDs are of three groups. As the board temperature increasing from 30°C to 120°, the first group performed monotonically decrease of CCT. On the contrary, the second group performed monotonically increase of CCT when the board temperature changes from 30° C to 120°C. Nevertheless, the third group of CCT self-compensation performed small CCT drifts as the board temperature changed. In the beginning, the red shift of the LED die enhances the phosphor excitation efficiency, and then results in decrease of CCT. Then, the heat flows to the phosphor and thermal quenching of the phosphor degraded the efficiency such as to turn CCT to higher value. Therefore, a turning point of each pcW-LED exists in the third group. In the experiment, we have successfully demonstrated to confine the deviation of CCT as small as 7 K for neutral white cases and 83 K for cool white cases. In the meantime, the new introduced G# factor is useful to the evaluation on whether the spectrum of the blue light matches to the excitation spectrum of the phosphor and the degree of CCT stabilization.
The proposed scheme is simple and useful in LED solid-state lighting with use of existing phosphors. The stabilizing CCT technology will benefit pcW-LEDs in not only general lighting but also other applications such as backlight for liquid crystal display.
Funding
MOE provided through National Central University’s Plan to Develop First-class Universities and Top-level Research Centers (103G-903-2); Ministry of Science and Technology of the Republic of China (MOST103-2221-E-008-063-MY3, MOST 104-2911-I-008-505 and MOST105-3113-E-008-008-CC2).
References and links
1. Y. Shimizu, “Development of White LED light source,” Rare Earths 40 (Rare Earth Society of Japan, 2002), pp. 150–151.
2. A. Zukauskas, M. S. Shur, and R. Caska, Introduction to Solid-state Lighting (John Wiley & Sons, 2002).
3. D. A. Steigerwald, J. C. Bhat, D. Collins, R. M. Fletcher, M. O. Holcomb, M. J. Ludowise, P. S. Martin, and S. L. Rudaz, “Illumination with solid state lighting technology,” IEEE J. Sel. Top. Quantum Electron. 8, 310–320 (2002).
4. E. F. Schubert and J. K. Kim, “Solid-state light sources getting smart,” Science 308(5726), 1274–1278 (2005). [PubMed]
5. C. C. Sun, Y. Y. Chang, T. H. Yang, T. Y. Chung, C. C. Chen, T. X. Lee, D. R. Li, C. Y. Lu, Z. Y. Ting, B. Glorieux, Y. C. Chen, K. Y. Lai, and C. Y. Liu, “Packaging efficiency in phosphor-converted white LEDs and its impact to the limit of luminous efficacy,” J. Solid State Lighting 1, 19 (2014).
6. C. Martinsons and G. Zissis, “Solid state lighting annex: potential health issues of solid state lighting,” Final Report, SSL Annex Task 1, Energy Efficient End-Use Equipment (4E) (IEA, 2014).
7. N. Narendran, Y. Gu, J. P. Freyssinier-Nova, and Y. Zhu, “Extracting phosphor-scattered photons to improve white LED efficiency,” Phy. Status Solidi A 202, 60–62 (2005).
8. T. Y. Chung, S. C. Chiou, Y. Y. Chang, C. C. Sun, T. H. Yang, and S. Y. Chen, “Study of temperature distribution within pcW-LEDs using the remote-dome phosphor package,” IEEE Photonics J. 7, 1–11 (2015).
9. B. J. Shih, S. C. Chiou, Y. H. Hsieh, C. C. Sun, T. H. Yang, S. Y. Chen, and T. Y. Chung, “Study of temperature distributions in pc-WLEDs with different phosphor packages,” Opt. Express 23(26), 33861–33869 (2015). [PubMed]
10. C. D. Thurmond, “The standard thermodynamic functions for the formation of electrons and holes in Ge, Si, GaAs, and GaP,” J. Electrochem. Soc. 122, 1133–1141 (1975).
11. J. Cho, C. Sone, Y. Park, and E. Yoon, “Measuring the junction temperature of III-nitride light emitting diodes using electro-luminescence shift,” Phys. Status Solidi., A Appl. Mater. Sci. 202, 1869–1873 (2005).
12. R. Mueller-Mach, G. O. Mueller, M. R. Krames, and T. Trottier, “High-power phosphor-converted light-emitting diodes based on III-nitrides,” IEEE J. Sel. Top. Quantum Electron. 8, 339–345 (2002).
13. S. Nakamura, “The roles of structural imperfections in InGaN-based blue light-emitting diodes and laser diodes,” Science 281(5379), 955–961 (1998). [PubMed]
14. P. Waltereit, O. Brandt, A. Trampert, H. T. Grahn, J. Menniger, M. Ramsteiner, M. Reiche, and K. H. Ploog, “Nitride semiconductors free of electrostatic fields for efficient white light-emitting diodes,” Nature 406(6798), 865–868 (2000). [PubMed]
15. Y. Zhao, C. Y. Huang, S. Tanaka, C. C. Pan, K. Fujito, D. Feezell, J. S. Speck, S. P. DenBaars, and S. Nakamura, “Blue and green InGaN light-emitting diode,” in CLEO/QELS 2012 (OSA, Baltimore, MD, USA, 2012), p. JTh4J.2.
16. M. E. Levinshteĭn, S. L. Rumyantsev, and M. Shur, Properties of Advanced Semiconductor Materials: GaN, AlN, InN, BN, SiC, SiGe (Wiley, 2001).
17. DOE SSL technology fact sheet, “Optical Safety of LEDs,” PNNL-SA-96340 (June 2013).
18. K. N. Kim, J. M. Kim, K. J. Choi, J. K. Park, and C. H. Kim, “Synthesis, characterization, and luminescent properties of CaS:Eu phosphor,” J. Am. Ceram. Soc. 89, 3413–3416 (2006).
19. H. S. Jang and D. Y. Jeon, “Yellow-emitting Sr3SiO5:Ce3+, Li+ phosphor for white-light-emitting diodes and yellow-light-emitting diodes,” Appl. Phys. Lett. 90, 041906 (2007).
20. J. Chen, J. Liu, and H. Yao, “Preparation and application of strong near-infrared emission phosphor Sr3SiO5:Ce3+, Al3+, Nd3+,” J. Am. Ceram. Soc. 98, 1836–1841 (2015).
21. R. J. Xie and H. T. Hintzen, “Optical properties of (oxy) nitride materials: a review,” J. Am. Ceram. Soc. 96, 665–687 (2013).
