Enhancement of luminous efficacy for LED lamps by introducing polyacrylonitrile electrospinning nanofiber film YONG TANG, ZHI LI, GUANWEI LIANG, ZONGTAO LI,* JIASHENG LI, AND BINHAI YU Key Laboratory of Surface Functional Structure Manufacturing of Guangdong High Education Institutes, School of Mechanical and Automotive Engineering, South China University of Technology, Guangdong, China *meztli@scut.edu.cn

Abstract: Polyacrylonitrile electrospinning nanofiber film was introduced into a light emitting diode (LED) lamp to exploit the strong reflective and scattering effects. The light extraction mechanism was studied systematically for three different electrospinning types in three different types of LED lamps. For the all-electrospinning types, the luminous efficacy increased for the white LED, outwards remote phosphor layer, and inwards remote phosphor layer lamps by 10.98, 16.97, and 18.35%, respectively, compared with the reference lamp. Lamps with stronger backscatter had larger luminous efficacy enhancements. The reflector-electrospinning type helped redirect lights with large emission angles. The substrate-electrospinning type was beneficial for recycling the total interior reflection lights and increasing the yellow to blue ratio. Additionally, the all-electrospinning white LED lamps remains 97.89% luminous flux after a 96-hour aging process. Electrospinning fiber films are favorable luminous efficacy enhancers for the future generation of LED lamps.

© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. Introduction Light emitting diodes (LEDs) based on GaN-based chips and color-converted elements have gained significant attention recently in the fields of general lighting, displays, indicators, and communication [1]. In the general lighting field, multiple LEDs and closed packages will be adapted to high-power-operation lamps to ensure reliability, and that lighting requirements are met [2]. Apart from inserting white LEDs into the modules, the combination of blue LEDs with remote phosphor layers (RPLs) produce a higher luminous efficacy and heat dissipation in the modules [3]. Other requirements, such as large area lighting without glare effects are desired in high-end applications, and an appropriate light diffusion plate is introduced for antiglare effects [4]. However, the closed package, RPLs, and light diffusion plate of the LED modules are associated with severe total interior reflection (TIR) effects. Most of the TIR lights are absorbed by the leadframe and substrate of the module, lowering the light extraction efficiency (LEE) and directly decreasing the luminous efficacy of the LED module [5].

Various approaches have been previously proposed to improve the LEE by reducing the TIR. At the chip level, both the DA series LED chip from CREE with V-grooves on a SiC substrate [6] and the chip fabricated by Mao et al. [7] with the PS/ZnO micro-nanostructure on top, are able to initially enhance LEE under a costly and complicated process. At the substrate level, a TiO2 reflecting layer [8] and a patterned leadframe substrate [9] were suggested for increasing the substrate reflection and recycling the TIR lights in the LED COB modules; however, the usage is limited and enhancement is insufficient. At the reflector cup level, bead blasting [10] and single-crystal silicon oxynitride nanowires [11] were introduced to enhance the scattering ability of the cup surface; however, the raw materials and production lines are expensive. At the remote structure design level, the precise lens design

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27716

#335839 https://doi.org/10.1364/OE.26.027716 Journal © 2018 Received 21 Jun 2018; revised 26 Sep 2018; accepted 27 Sep 2018; published 9 Oct 2018

[12] and RPL design [13] are useful in matching the original extracted lights in each emission angle and lower the TIR directly; however, different modules require different structure designs, which is time consuming.

As a simple and efficient method to fabricate nanofibers (NFs), electrically assisted spinning (electrospinning) technology has recently received considerable interest in many industries [14], including the optics [15], electrics [16], biomedical [17], and mechanical [18] fields. By simply controlling the simple flow rate and concentration, identical and ideal NF films can be obtained with desired physical and chemical properties [19], which is simple and valuable for mass production. Owing to the nano-size of the fibers and the staggered film structure, the electrospinning NF films possess strong light scattering and reflectance abilities, attracting interest in many research fields. Mie theory was introduced to analyze the scattering ability of electrospinning NF films, and significant scattering in the visible wavelength was demonstrated [20]. Chang et al. and our previous research demonstrated that the NF diameters and film thicknesses have an approximately positive linear relation with the scattering ability [21,22]. Moreover, Lee et al. covered electrospinning NF films on top of LED and organic LED modules to control the scatter of extraction lights [4]. Our previous study proved that a great angular color uniformity enhancement could be achieved though electrospinning poly(lactic-co-glycolic acid) NF films on the reflector surface of the LED modules [22]. Ye et al. prepared a hollow fibrous film with high reflectance by coaxial electrospinning and pointed out their light enhanced potential in backlight area [23]. Luminous efficacy is the most essential performance index for the LED modules in general lighting, however, the research of utilizing the strong scatter and reflectance ability of the electrospinning NF films to influence the luminous efficacy of the LED modules is still scarce.

In this study, polyacrylonitrile (PAN) NF films with different thicknesses were first fabricated by the electrospinning process. The reflectance of each of the NF films was tested and the one with the highest value was introduced into an experimental LED lamp. With none-E, R-E, S-E, and A-E types; the radiant fluxes of blue LED lamps, luminous efficacy of white LED lamps, outwards remote phosphor layer (ORPL) lamps, and inwards remote phosphor layer (IRPL) lamps were studied systematically. The reliability of the lamp was study. The luminous efficacy enhanced mechanism of the NF films was revealed, providing an effective solution for increasing the luminous efficacy of the LED modules in the general lighting field.

2. Experiment 2.1 Preparation of electrospinning films

The PAN (with a molecular weight of 80000 g/mol, Aladdin) was first mixed in dimethylformamide (DMF, Richjoint) solvent with a mass-volume concentration of 15%. After 6 h of magnetic stirring under the ambient temperature, the PAN polymer solution stood for 12 h for the removal of the air bubbles. As shown in Fig. 1, the prepared PAN polymer solution was inserted into a syringe that was fastened to the Lange syringe pump (model LSP01-1A, manufactured by Baoding Lange Constant Flow Pump Co., Ltd.). Aluminum foil was covered tightly over the squared board to serve as the electrospinning fiber deposition substrate, and the distance between the syringe tip and substrate was 14 cm. A 15 kV high voltage electric field and 0.4 ml/h injection speed were provided by a high-voltage power supply (model DW-P501-1ACDF, supplied by Tianjin Dongwen High Voltage Power Supply Co., Ltd.) and the pump respectively. The PAN NFs could be pulled out from the syringe tip (positive pole) and deposited onto the aluminum substrate (negative pole) to form the NF films.

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27717

2.2 Preparation of lamps

The closed lamp consisted of an aluminum shell, a circuit substrate with a white oil coating, an isolated polymer reflector cup, a diffusion plate, and a fixed frame. Three blue or three white LEDs (3535 semi-sphere lens package, 448-nm blue chip, provided by Foshan National Star Co., Ltd.) were mounted onto the circuit substrate. Using the adhesive layer, the NF films can be transferred and adhered precisely and flat onto the circuit substrate and reflector cup. The areas of the NF films adhering to the substrate and reflector cup were approximately 2460 and 1760 mm2, respectively.

Fig. 1. Schematic of the electrospinning process and lamp.

2.3 Preparation of remote phosphor layer

To fabricate the RPL, SDY558 yellow phosphor (bought from Yantai Blight Photoelectric Material Co., Ltd.), R615 red phosphor (bought from Shanghai Elumin Innovative Materials Co., Ltd.), and polydimethylsiloxane (PDMS, bought from Dow Corning Sylgard-184) with a mass ratio of 1:0.03:5.04 was mixed first. The mass ratio was the same as that used in the white LEDs. Then, 0.8 ml of the prepared phosphor mixture was coated directly onto the diffusion plate and spun under 500 rpm. After the spin-coating process, the uniform and flat RPL was cured at 110 °C for 30 min before used.

2.4 Measurement and testing

In the experiment, the morphology of the PAN electrospinning NF film was characterized by a high-resolution scanning electron microscope (SEM, Zeiss Merlin) under an accelerated voltage of 5.0 kV. The diameter of the PAN electrospinning NF was measured using the ImageJ package and 20 NFs were measured to determine the average fiber diameter. Moreover, a surface profilometer (Vecco Dektak 150, America) was introduced to measure the thickness of the PAN electrospinning NF films.

An ultraviolet spectrophotometer (dual-beam UV-Vis Spectrophotometer TU-1901) was used to characterize the reflectance property of each diffuse element. Moreover, the light intensity distribution was tested using a spatial light distribution instrument [22] under 450-nm to represent the light scattering ability of each element. The photoelectric test of each lamp was conducted by a calibrated integrating sphere under a driving current from 75 to 750 mA.

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27718

3. Results and discussion 3.1 Characteristics of the PAN electrospinning NF films

The thickness and reflectance of the PAN electrospinning NF films are shown in Fig. 2(a). As the deposition time increased, the PAN NFs extracted from the syringe tip were stacked layer by layer onto the substrate, and the thickness of the PAN electrospinning NF films increased from 64.7 μm (at 15 min) to 212.6 μm (at 120 min). The thicker and denser structure directly gave rise to a reflectance increase of the PAN electrospinning NF films; the reflectance increased from 88.69% (at 15 min) to 98.20% (at 120 min) at 450 nm, which were significantly higher than those of the origin white oil coating circuit substrate (82.81%) and polymer reflector cup (85.42%) of the lamp. Figure. 2(b) shows the light intensity distribution of the substrate, reflector, and PAN electrospinning NF film for a 120-min deposition time. As illustrated, the light intensity distribution of the PAN electrospinning NF film was much more scattered compared with the reflector and substrate. Hence, it was confirmed that the PAN electrospinning NF film provides an effective scattering ability. Figure. 2(c) shows a high-resolution SEM image of the 120-min deposition time PAN electrospinning NF film chosen for further study. The NFs with uniform diameters of 220 nm showed unordered stacking and formed a dense multiple layer.

Fig. 2. (a) Film thickness and reflectance (450 nm) of the PAN electrospinning NF films with different deposition times; the red and green dashed lines represent the reflectances (450 nm) of the substrate and reflector, respectively. (b) Light intensity distribution of the substrate, reflector, and PAN electrospinning NF film for a 120-min deposition time. (c) High resolution SEM image of the 120-min-deposition-time PAN electrospinning NF film.

3.2 Overall optical performance of the LED lamps with PAN electrospinning NF films

To simply specify the impact of the high reflectance PAN electrospinning NF films on the LED lamp, we chose the 120-min-deposition-time sample with the highest reflectance of 98.20% to precisely adhere to the circuit substrate and polymer reflector. Different types of experimental lamps were studied: a lamp with no PAN electrospinning NF film (for reference), a lamp with PAN electrospinning NF film adhered to the reflector (reflector-

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27719

electrospinning, R-E), a lamp with PAN electrospinning NF film adhered on substrate (substrate-electrospinning, S-E) and a lamp with PAN electrospinning NF film adhered to both the reflector and substrate (all-electrospinning, A-E). The blue LED lamps without phosphor conversion were studied first, and Fig. 3(a) shows the radiant flux of these four electrospinning types of blue LED lamps. The radiant flux of each lamp increased consistently with the rising input current. And it indicates that the PAN electrospinning NF film provides a significant increase in the radiant flux of each type of electrospinning-introduced lamp, compared with the reference lamp under every driving current. Specifically, the A-E lamp had the highest radiant flux among all of the types of blue LED lamps, and the radiant flux of the S-E lamp was larger than that of the R-E lamp. As shown in the inset of Fig. 3(a). The average radiant flux increases of the A-E lamp, S-E lamp, and R-E lamp were 13.15, 8.90, and 5.76% respectively. By increasing the reflectance of the lamp cavity, the TIR lights had a greater chance of being redirected by the PAN electrospinning films and thus escape with an absorption loss decrease.

We next studied cases with phosphor conversion. There are two widely used methods for obtaining white light lamps [24,25]. The first is to directly mount white LEDs into the lamp, which was the method used in this experiment for the white LED lamp. The second method is to mount blue LEDs into the lamp and the blue light excites the RPL to produce white light, which corresponds to the outwards RPL (ORPL, with the RPL facing outwards) and inwards RPL (IRPL, the RPL faces towards the substrate) lamps used in this experiment.

Fig. 3. (a) Radiant flux of four types of blue LED lamps under a driving current from 75 to 750 mA; the inset shows the average radiant flux increases of the R-ES, S-ES, and A-ES lamps, the inset shows the average radiant flux increase compared to the reference lamp. The luminous efficacy of the electrospinning-introduced to the (b) white light LED, (c) ORPL, and (d) IRPL lamps under a driving current from 75 to 750 mA.

As can be seen from the Figs. 3(b) - (d), the luminous flux of each type of white LED, ORPL, and IRPL lamps increased consistently with the rising input current, while the luminous efficacy decreased. The A-E lamps produced the highest luminous efficacy, and the luminous efficacy of the S-E lamps was larger than that of the R-E lamps. These result show

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27720

that the luminous performance of the PAN electrospinning NF film lamps clearly surpassed the reference lamps. To distinguish the luminous efficacy variance of each lamp, we listed the correlated color temperature (CCT), luminous efficacy, and luminous efficacy increased at a driving current of 300 mA compared to the reference lamp in the Table 1.

Table 1. CCT, luminous efficacy, and luminous efficacy increase compared to the reference lamp for the electrospinning-introduced white LED, ORPL, and IRPL lamps at

a 300 mA driving current.

Electrospinning type

CCT (K) / CCT variance (%)

Luminous efficacy (lm/W)

Luminous efficacy increase (%)

White LED lamp

Reference 4642 / - 126.00 - R-E 4639 / 0.06 130.44 3.52 S-E 4583 / 1.27 136.39 8.25 A-E 4577 / 1.40 139.84 10.98

ORPL lamp

Reference 4590 / - 135.73 - R-E 4567 / 0.50 142.01 4.63 S-E 4457 / 2.90 152.36 12.25 A-E 4443 / 3.20 158.76 16.97

IRPL lamp

Reference 4410 / - 138.34 - R-E 4402 / 0.18 146.60 5.97 S-E 4271 / 3.15 158.02 14.23 A-E 4267 / 3.24 163.73 18.35

As shown in Table 1, the controlled CCT between the white LED, ORPL, and IRPL lamps under the same electrospinning type had a variance of ~200 K. With the same phosphor ratio, a smaller CCT implies a higher blue light conversion ability of the lamp. Introducing PAN electrospinning NF film into the lamp slightly reduced the CCT of each type of lamp, and the maximum CCT variance of different electrospinning types for the white LED, ORPL, and IRPL lamps was 1.40, 3.20, and 3.24%, respectively. Under the relatively small CCT variance, the luminous efficacy of each type of lamps showed a significant increase compared with reference lamp. Of all the lamps, the A-E electrospinning type showed the highest luminous efficacy enhancement, followed by the S-E and R-E electrospinning types in order. For all of the electrospinning types, the statistics demonstrate that the IRPL lamp benefited the most from the introduction of the PAN electrospinning NF film: the A-ES IRPL lamp showed a 18.35% luminous efficacy increase, compared those of the A-ES ORPL and A-ES white LED lamps of 16.97 and 10.98%, respectively.

3.3 Optical mechanism study of the LED lamps with PAN electrospinning NF film

To further study the optical performance of the three types of lamps with four electrospinning types, the radiant flux is shown in Fig. 4(a). The A-E structure has the highest radiant flux among all of the electrospinning types in the three types of lamps, demonstrating that the introduction of electrospinning NF films in the LED lamp significantly enhances the radiant flux, which is essential for lamp applications. Compared to the reference lamp, the radiant flux enhancement of the A-E IRPL lamp was the highest (14.59%), followed by the A-E ORPL lamp (13.84%) and A-E white LED lamp (10.41%) in order. The increase of the luminous efficacy did not sacrifice the radiant power, and the radiant flux enhancement of the different types of lamps varied. In particular, the RPL lamps, especially the IRPL lamp, had a higher radiant flux enhancement compared with the white LED lamp. As phosphor contains a stronger backscattering ability [26], the TIR phenomenon was more severe in RPL lamps, especially the IRPL lamp. The high scattering and reflectance abilities of the PAN NF films in the lamp cavity helped to redirect and extract those TIR lights more significantly, and therefore the lamps with stronger backscattering benefited more significantly.

The emission spectra of the three types of lamps are shown in Figs. 5(a) - 5(c). Firstly, it is obvious that the blue-yellow light conversion ability differed between lamps. The blue-yellow light conversion ability of the RPL lamps was larger than that of the white LED

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27721

lamps, which is mainly due to the higher phosphor conversion efficacy of RPL structures [27]. As the backscattering of the IRPL was stronger than that of ORPL, more blue lights were reflected into the lamp cavity, and more yellow light was converted, and thus a higher blue-yellow conversion ratio could be achieved. According to the definition of the visual function [28], a higher yellow light proportion brings a higher luminous efficacy, which is in accordance with the results in Table 1.

Furthermore, compared with that of the non-electrospinning reference lamp, the light intensity of the three types of electrospinning introduced lamps had a significant increase over the whole wavelength range. As for the yellow light intensity, the A-E lamps had the highest value for all three types of lamps, followed by the S-E and R-E lamps in order, which was also observed for the luminous efficacy. However, this was not the case when ranking the blue light intensity results. As the A-E lamps had the highest blue light intensity due to the highest scatter and reflectance in the lamp cavity, the R-E lamps had a higher blue light intensity compared with the S-E lamps in the RPL lamps.

Fig. 4. Radiant flux of three types of lamps under a 300 mA driving current;

Fig. 5. Emission spectra of the four electrospinning types of (a) white LED, (b) ORPL, and (c) IRPL lamps under a 300-mA driving current.

To further investigate this issue, the blue light radiant flux at 375–480 nm and yellow light radiant flux at 480–800 nm were individually extracted by integrating the corresponding energy of the spectra. The blue and yellow light radiant flux increase ratios of each PAN electrospinning lamp compared with the reference lamp are shown in Fig. 6(a). In general, the yellow light increase ratio was larger than that of the blue light for each electrospinning type because the increase of the yellow light benefited from both the reduction of the total internal reflection and higher phosphor conversion. Consequently, the yellow light increase, which is more sensitive to the human eye, led to a positive influence on the luminous efficacy increase of the lamps. Both the blue and yellow light increases for the R-E type showed little difference among the three types of lamps, revealing that the different backscattering abilities of the lamps made little difference to the light promotion mechanism. The R-E structure mainly promoted the initial extracted blue/yellow light with a large emission angle, and the redirected blue/yellow light had the potential of being converted by the RPL or directly

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27722

emitted with a small angle. As for the S-E type, the blue light increase of the RPL lamps was significantly smaller than that of the white LED lamp, demonstrating that a portion of the TIR blue light was reflected by the S-E type then absorbed and transferred by the RPL, and therefore, the yellow light increase was consequently larger. The IRPL lamp with a higher backscattering ability benefited the most from the S-E type. The A-E type combines the characteristics of the R-E and S-E types, and the blue and yellow increases of each lamps are maximized.

To further investigate this issue, the yellow-to-blue ratio (YBR) of three types of lamps are shown in Fig. 6(b). Firstly, the IRPL lamp had a higher initial YBR because the light conversion is higher in the lamp cavity due to the stronger backscatter. By adding the R-E structure, the blue light with a large emission angle of the RPL lamps have the chance to be extracted, and are converted into yellow light with a similar proportion as that of the reference lamps; therefore, the YBR of the RPL lamps is almost constant in the R-E structure. While the lambertian blue lights with the large emission angle of the white LED lamp had a higher extraction increase, compared with the isotropically emitted yellow lights, the YBR of the white LED lamp decreased slightly in the R-E structure. For the S-E structure, the YBR of the three types of lamps showed a significant increase, and it can be seen from the slope that the increase range of the IRPL lamp was larger than that of the ORPL lamp, which in turn was larger than that of the white LED lamp. As the PAN electrospinning NF films possess a strong reflectance and scattering ability, almost no TIR blue light was absorbed by the lamp cavity. Hence, stronger backscattering provides a more sufficient absorption of blue light by the phosphor, and thus a higher YBR is obtained. The YBR of the A-E structure is obtained by combining the characteristics of the R-E and S-E types.

Fig. 6. (a) Blue and yellow light increase ratios of three types of lamps. (b) YBR of three types of lamps.

Finally, we had conducted an aging experiment selecting the white LED lamp (the lamp that most commonly used in the market) of all-electrospinning type for 96 hours. As illustrated in the Fig. 7, the luminous flux of the A-E white LED lamp remained 97.89% after a 96-hour aging process, which was comparable to the commonly used LEDs. The aging performance of the A-ES white LED lamp proves the stable property of NF films used in the general lighting.

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27723

Fig. 7. Luminous flux change of the A-E white LED lamp under a 96-hour aging process; inset shows the top view of the lamp under operation.

The results can be summarized as follows: 1) The introduction of high reflectance PAN electrospinning NF films into the lamps significantly enhanced the radiant flux and luminous efficacy simultaneously, and the more severe the backscatter in the lamp, the greater the enhancement. 2) The R-E structure was beneficial for improving the light extraction for large emission angles, while the S-E structure is useful for recycling the TIR lights and increasing the YBR. 3) When using the A-E structure, the radiant power and luminous efficacy of each lamps are maximized.

4. Conclusion A PAN electrospinning NF film with 98.20% reflectance was fabricated and introduced into LED lamps, and the relative light extraction of white LED, ORPL, and IRPL lamps were studied systematically. The blue light lamp was used to demonstrate that R-E, S-E, and A-E structures could be used to enhance the blue light extraction to varying degrees, owing to the strong scattering and reflective abilities of the film. The luminous efficacy of these three types of lamps were studied as well. The luminous efficacy of the IRPL lamp with an A-E structure increased by 18.35%, followed by increases of 16.97 and 10.98% for the corresponding ORPL and white LED lamps, respectively (300 mA). Upon introduction of the electrospinning NF films, the more severe the backscatter of the lamp, the higher the radiant power and luminous efficacy enhancement. By analyzing the blue and yellow light variations separately, it was found that the R-E structure can improve the light extraction at large emission angles, while the S-E structure was beneficial for recycling the TIR lights and increasing the YBR. While using the A-E structure, the radiant flux and luminous efficacy of each lamps were maximized. Additionally, the A-ES white LED lamp remains 97.89% luminous flux after a 96-hour aging process. We believe that the high reflectance and scattering of the electrospinning NF film will significantly contribute to the development of LED lamp applications.

Funding National Natural Science Foundation of China (NSFC) (51775199, 51735004); Natural Science Foundation of Guangdong Province (2014A030312017); Science and Technology Program of Guangdong Province (2016B010130001).

References and links 1. J. Cho, E. F. Schubert, and J. K. Kim, “Efficiency droop in light‐emitting diodes: challenges and

countermeasures,” Laser Photonics Rev. 7(3), 408–421 (2013). 2. B. D’Andrade, “Lighting: White phosphorescent LEDs offer efficient answer,” Nat. Photonics 1(1), 33–34

(2007).

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27724

3. Z. T. Li, H. Y. Wang, B. H. Yu, X. R. Ding, and Y. Tang, “High-efficiency LED COB device combined diced V-shaped pattern and remote phosphor,” Chin. Opt. Lett. 15(4), 56–59 (2017).

4. H. J. Lee, S. An, J. H. Hwang, S. G. Jung, H. S. Jo, K. N. Kim, Y. S. Shim, C. H. Park, S. S. Yoon, Y. W. Park, and B. K. Ju, “Novel composite layer based on electrospun polymer nanofibers for efficient light scattering,” ACS Appl. Mater. Interfaces 7(1), 68–74 (2015).

5. M. H. Shin, H. G. Hong, H. J. Kim, and Y. J. Kim, “Enhancement of optical extraction efficiency in white LED package with quantum dot phosphors and air-gap structure,” Appl. Phys. Express 7(5), 052101 (2014).

6. X. Ding, Y. Tang, Z. Li, J. Li, Y. Xie, and L. Lin, “Multichip LED modules with V-Groove surfaces for light extraction efficiency enhancements considering roughness scattering,” IEEE Trans. Electron Dev. 64(1), 182–188 (2017).

7. P. Mao, A. Mahapatra, J. Chen, M. Chen, G. Wang, and M. Han, “Fabrication of polystyrene/ZnO micronano hierarchical structure applied for light extraction of light-emitting devices,” ACS Appl. Mater. Interfaces 7(34), 19179–19188 (2015).

8. H. Zheng, L. Li, X. Lei, X. Yu, S. Liu, and X. Luo, “Optical performance enhancement for chip-on-board packaging LEDs by adding TiO2/silicone encapsulation layer,” IEEE Electron Device Lett. 35(10), 1046–1048 (2014).

9. Z. T. Li, Q. H. Wang, Y. Tang, C. Li, X. R. Ding, and Z. H. He, “Light extraction improvement for LED COB devices by introducing a patterned leadframe substrate configuration,” IEEE Trans. Electron Dev. 60(4), 1397–1403 (2013).

10. J. K. Kim, H. Luo, E. F. Schubert, J. Cho, C. Sone, and Y. Park, “Strongly enhanced phosphor efficiency in GaInN white light-emitting diodes using remote phosphor configuration and diffuse reflector cup,” Jpn. J. Appl. Phys. 44(20–23), L649–L651 (2014).

11. S. Xi, T. Shi, L. Zhang, D. Liu, W. Lai, and Z. Tang, “Highly visible-light reflective SiO x N y nanowires for bright-white reflector applications,” Thin Solid Films 529(2), 115–118 (2013).

12. Z. Li, S. Yu, L. Lin, Y. Tang, X. Ding, W. Yuan, and B. Yu, “Energy feedback freeform lenses for uniform illumination of extended light source LEDs,” Appl. Opt. 55(36), 10375–10381 (2016).

13. J. Li, Z. Li, Z. Li, Y. Tang, X. Ding, and B. Yu, “Improvement in optical performance and color uniformity by optimizing the remote phosphor caps geometry for chip-on-board light emitting diodes,” Solid-State Electron. 126, 36–45 (2016).

14. N. Bhardwaj and S. C. Kundu, “Electrospinning: A fascinating fiber fabrication technique,” Biotechnol. Adv. 28(3), 325–347 (2010).

15. R. Dersch, M. Steinhart, U. Boudriot, A. Greiner, and J. H. Wendorff, “Nanoprocessing of polymers: applications in medicine, sensors, catalysis, photonics,” Polym. Adv. Technol. 16(2–3), 276–282 (2010).

16. D. Wu, C. Shi, S. Huang, X. Qiu, H. Wang, Z. Zhan, P. Zhang, J. Zhao, D. Sun, and L. Lin, “Electrospun nanofibers for sandwiched polyimide/poly (vinylidene fluoride)/polyimide separators with the thermal shutdown function,” Electrochim. Acta 176, 727–734 (2015).

17. A. Periyakaruppan, P. U. Arumugam, M. Meyyappan, and J. E. Koehne, “Detection of ricin using a carbon nanofiber based biosensor,” Biosens. Bioelectron. 28(1), 428–433 (2011).

18. C. H. Hung and W. F. Leung, “Filtration of nano-aerosol using nanofiber filter under low Peclet number and transitional flow regime,” Separ. Purif. Tech. 79(1), 34–42 (2011).

19. M. M. Munir, F. Iskandar, K. M. Yun, K. Okuyama, and M. Abdullah, “Optical and electrical properties of indium tin oxide nanofibers prepared by electrospinning,” Nanotechnology 19(14), 145603 (2008).

20. Y. L. Chen, Y. H. Chang, J. L. Huang, I. Chen, and C. Kuo, “Light scattering and enhanced photoactivities of electrospun titania nanofibers,” J. Phys. Chem. C 116(5), 3857–3865 (2012).

21. C. C. Chang, C. M. Huang, Y. H. Chang, and C. Kuo, “Enhancement of light scattering and photoluminescence in electrospun polymer nanofibers,” Opt. Express 18(S2 Suppl 2), A174–A184 (2010).

22. Y. Tang, G. Liang, J. Chen, S. Yu, Z. Li, L. Rao, and B. Yu, “Highly reflective nanofiber films based on electrospinning and their application on color uniformity and luminous efficacy improvement of white light-emitting diodes,” Opt. Express 25(17), 20598–20611 (2017).

23. C. Ye, M. Li, J. Hu, Q. Cheng, L. Jiang, and Y. Song, “Highly reflective superhydrophobic white coating inspired by poplar leaf hairs toward an effective “cool roof”,” Energy Environ. Sci. 4(9), 3364–3367 (2011).

24. H. C. Chen, J. Y. Lin, and H. Y. Chiu, “Rectangular illumination using a secondary optics with cylindrical lens for LED street light,” Opt. Express 21(3), 3201–3212 (2013).

25. H. Luo, K. S. Kim, Y. A. Xi, E. Fred Schubert, J. Cho, C. Sone, and Y. J. Park, “Trapped whispering-gallery optical modes in white light-emitting diode lamps with remote phosphor,” Appl. Phys. Lett. 89(4), 216–221 (2006).

26. J. Li, Y. Tang, Z. Li, X. Ding, L. Rao, and B. Yu, “Effect of quantum dot scattering and absorption on the optical performance of white light-emitting diodes,” IEEE Trans. Electron Dev. 65 (7), 2877–2884 (2018).

27. S. P. Ying and H. Y. Chien, “Effect of reassembled remote phosphor geometry on the luminous efficiency and spectra of white light-emitting diodes with excellent color rendering property,” IEEE Trans. Electron Dev. 63(3), 1117–1121 (2016).

28. Y. Tang, Z. Li, Z. T. Li, J. S. Li, S. D. Yu, and L. S. Rao, “Enhancement of luminous efficiency and uniformity of CCT for quantum dot-converted LEDs by incorporating with ZnO nanoparticles,” IEEE Trans. Electron Dev. 65(1), 158–164 (2017).

Vol. 26, No. 21 | 15 Oct 2018 | OPTICS EXPRESS 27725