Improved color rendering index of low band gap semi-transparent polymer solar cells using one-dimensional photonic crystals

Abstract

Neutral color semi-transparent polymer solar cells (STPSCs) have attracted a lot of attention due to their unique application in building integrated photovoltaic (BIPV) in recent years. Herein, we report a type of low band gap STPSCs with high color rendering index (CRI), which can be modulated by one-dimensional photonic crystals (1DPCs). To flatten transmitted light and eliminate chromatic aberration, the 1DPCs were integrated on top of the STPSCs as a wavelength-dependent filter. We have designed the centre wavelength of 1DPC at 520 nm to level the concavo-convex transmittance spectrum induced by the absorption of PSBTBT:PC₆₀BM. By turning the pairs of 1DPCs, the optimal STPSCs with 3 pairs of 1DPC achieved an increased CRI from 77 to 91 under an AM 1.5G illumination light source, a suppressed chromaticity difference (DC) from 0.0308 to 0.0045 and a enhanced power conversion efficiency (PCE) from 3.43% to 4.01% compared to devices without 1DPC. Furthermore, we believe that this new method for STPSCs incorporating 1DPCs filters to balance the total transmittance spectrum and enhance the CRI will play a important role in future BIPV applications.

---

1. Introduction

As we all know, solar energy is a renewable, safe and powerful energy source.^1,2 One of the most effective ways to utilize it is using solar cells, which can directly convert solar light into electricity.³ However, the majority of solar energy suffers great losses in the ocean, ground and surface infrastructure instead of being collected for utilization, particularly in buildings, walls and automotive skin used in urban life. Therefore, a novel application of solar energy has emerged in recent years, that is building integrated photovoltaic (BIPV), which combine buildings with huge photovoltaic applications.^4–6 To meet the requirements of keeping light incidence and supplying power simultaneously for building windows, semi-transparent polymer solar cells (STPSCs), which integrate optical and electrical characteristics in one, are regarded as a potential candidate for BIPV applications.^7–11

In the past few decades, investigations towards STPSCs mainly have focused on designing different light trapping structures and synthesizing various polymer materials to solve the trade-off between the efficiency and transmittance, simultaneously.^12,13 Moreover, to keep the transmitted light a neutral color is a critical factor for the purpose of BIPV window applications to meet the need for people's daily work and life.^14,15 It requires that the light passing through the STPSCs window has to keep the neutral color property of the incident light source such as the sun. In 2010, the color characterization of STPSCs was firstly studied as a milestone though it only simply depicted the color coordinate in color space.¹⁶ Then, the color rendering index (CRI) of STPSCs, which is a quantitative parameter of the neutral color property, began to be mentioned.^17–20 It is well known that the CRI is a quantitative measure of the ability of a light source to reveal the colors of various objects faithfully in comparison with an ideal light source, which is defined by the Commission Internationale de L'Eclairage (CIE, in French).^21–23 Light sources with a high CRI are desirable in color-critical applications such as neonatal care, photography and cinematography. The CRI value represents the approaching degree of the light source to neutral light. It is illustrated in the transmitted light quality in the STPSCs field. Note that the CRI is considered to be meaningful only if the color coordinates lie within a chromaticity difference (DC) < 0.0054 from the Planckian locus on the CIE 1960 uniform color space as recommended by the CIE. In terms of the CRI definition of CIE, the temperature of the nearest point of Planckian locus represents the correlated color temperature (CCT) and the distance between device color coordinate and CCT point on Planckian locus is the DC. The CRI is calculated by comparing the transmitted light with a reference black body radiator (BBR) with the same CCT on the Planckian locus. Referring to the CRI definition, the calculated results of CRI will become less accurate if the DC is larger than 0.0054 because the reference BBR with same CCT is not near to the color coordinate of the transmitted light of the STPSC. In order to achieve a higher CRI, some strategies have been reported, such as the utilization of a low band gap absorber,¹⁷ a tandem structure with complimentary absorbers,¹⁶ and a combination with dyes in the STPSCs.¹⁹ However, there are only a few reports with a CRI >90,^19,20 which can only be achieved by sacrificing efficiency. In addition, the DC is still higher than the tolerance of the CRI calculation.¹⁹ To date, how to achieve an STPSC with simultaneously improved CRI and PCE combined with suppressed DC is a challenge for BIPV applications.

In this study, one-dimensional photonic crystals (1DPCs) as a wavelength-dependent filter and light compensating mirror were integrated on top of a PSBTBT:PC₆₀BM-based²⁴ low band gap STPSC to flatten the transmitted light and eliminate chromatic aberration. By optimizing the period of 1DPC, a high CRI of 91 and a low DC of 0.0045 are achieved combined with an improved PCE from 3.43% to 4.01%. The centre wavelength of the 1DPCs was set at 520 nm to flatten the concavo-convex transmittance spectrum induced by the absorption of PSBTBT:PC₆₀BM, which lies between 600–800 nm. For the CRI and DC computation, the transmitted light is obtained by multiplying the transmittance with the AM 1.5G illumination light source spectrum. The all-round improved performance indicates that the method can balance the total device transmittance spectrum and enhance the color rendering properties of the CRI with favourable efficiency.

2. Experimental details

As shown in Fig. 1, the device structure is ITO/TiO₂/PSBTBT:PC₆₀BM/WO₃(10 nm)/Ag (15 nm)/1DPC ([WO₃(62.5 nm)/LiF(90.3 nm)]^pairs). First, the ITO glass substrate was cleaned with acetone, ethanol and de-ionized water for 30 minutes. Then, TiO₂ was prepared as the electron transport layer using sol–gel methods.²⁵ PSBTBT (Lumtec Corp) and PC₆₀BM (Lumtec Corp) were dissolved in 1,2-dichlorobenzene to produce a 15 mg ml⁻¹ solution in a 1 : 1 weight ratio and stirred for two days in a glove box. The thin film PSBTBT:PC₆₀BM (80 nm) was prepared by spin coating at 2000 rpm for 30 s and annealed at 110 °C for 30 minutes in a glove box. Then, 10 nm WO₃ and 15 nm Ag were deposited in sequence under 5 × 10⁻⁴ Pa vacuum. In order to determine the active area of the STPSC, the Ag top electrode was evaporated through a shadow mask. The dimension of the STPSC was 0.064 cm².

Fig. 1 Device structure of the STPSC with 1DPC.

1DPC is composed of alternately deposited layers of WO₃ and LiF.^26–28 We define the thickness and refractive index of WO₃ and LiF as d₁, d₂, n₁ and n₂. There is a relationship between n₁, n₂, d₁ and d₂, that is n₁d₁ = n₂d₂ = λ₀/4. λ₀ is the center wavelength of the 1DPC's photonic band gap, which is set at λ₀ = 520 nm. By measuring, the following parameters of n₁ = 2.08, n₂ = 1.44, d₁ = 62.5 nm and d₂ = 90.3 nm were obtained. The different pairs of [WO₃/LiF]^N (N = 2, 3, 4)²⁹ were alternately evaporated under vacuum (5 × 10⁻⁴ Pa) and the rate of deposition was about 1 nm s⁻¹. It can be seen that the overlap between ITO and Ag is regarded as the active layer and the 1DPCs were evaporated through a mask, which can cover the whole active layer well.

The absorption, transmittance and reflectance spectra were measured using an ultraviolet/visible spectrometer (UV1700, Shimadzu). J–V characteristics were measured using a Keithley 2601 source meter under an Oriel 300 W solar simulator with an intensity of 100 mW cm⁻² using a AM 1.5G illumination light source. The light source intensity was measured by a photometer (International light, IL1400). The incident photon-to-electron conversion efficiency (IPCE) was measured using a Crowntech QTest Station 1000AD. The calculation of the device color perception was done using Spectra Win software compiled by our team.

3. Results and discussion

3.1 Device color perception

The reflectance spectra of 1DPCs and the absorption spectrum of PSBTBT:PC₆₀BM are shown in Fig. 2. It can be seen that the absorption of PSBTBT:PC₆₀BM mainly covers the range between 600–800 nm due to the smaller band gap, which results in the case that the STPSCs based on PSBTBT:PC₆₀BM always appear as a blue-green color. In order to tune the specific color of the STPSCs into a neutral color and satisfy the application requirements, we have designed the 1DPCs with a complementary reflectance spectrum to flatten the transmittance spectrum of the STPSC. Here, we set 520 nm as the λ₀ value. It is apparent that the reflectance of 1DPC below 600 nm is increased gradually. It suggests that the 1DPCs can selectively modulate the light transmission as a filter according to the absorption range.

Fig. 2 Reflectance spectra of the 1DPCs corresponding to the left axis and the absorption spectrum of PSBTBT:PC₆₀BM with gray line corresponding to right axis.

In order to testify the function of the 1DPCs, the total device transmittance spectra of the respective STPSCs with different pairs of 1DPCs were measured and are illustrated in Fig. 3. For the control device without the 1DPCs, the shape of the transmittance spectrum is concavo-convex. The transmittance below 600 nm was obviously higher than that over 600 nm caused by the absorption spectrum. However, the shapes of the spectra curves were changed remarkably with the different pairs of 1DPCs, which show tuneable transparency. In perceptual intuition, the tall peak of the transmittance curve (around 480 nm) for the STPSC without the 1DPCs was shifted and descends along with the increasing pairs of 1DPCs induced by the higher reflectance. The amplitude of the transmittance spectra gradually decreases and gets more uniform. Such spectra suggest that the chromatic aberration of the transmitted light was decreased and the transmitted monochromatic light was more balanced. The highest CRI was obtained in the device with 3 pairs of 1DPCs and the compared photograph of the device with or without 3 pairs of 1DPCs in the real photo background is shown in the inset of Fig. 3. The background color seen through the STPSC was almost consistent except only a decrease in brightness. This was direct evidence that the 1DPCs can modulate the transmittance and improve the CRI effectively.

Fig. 3 Total transmittance spectra of the different devices with 0, 2, 3, 4 pairs of 1DPCs. Inside top left corner: a photograph of the device with 3 pairs of 1DPCs with a real photo background.

To analyze the device transparency color perception quantitatively, we calculated the color coordinates (x, y) on the CIE 1931 color space and (u, v) on the CIE 1960 uniform color space from the transmittance spectra, which are shown in Fig. 4. In the calculation, the light source was a AM 1.5G illumination light source. The transmitted light is represented by the product of the transmittance spectra of each STPSC and AM 1.5G illumination light source. CIE 1931 is the initial CIE color space, which first defined quantitative links between physical pure colors in the electromagnetic visible spectrum and physiological perceived colors in human color vision. This is the basis of color study. However, in the CIE 1931 color space, the same distance of color coordinate does not means the same color variation. In order to see the color changing intuitively, the CIE 1960 color space is useful, which is also named as the CIE 1960 uniform color space.

Fig. 4 (a) The representation of the color coordinates of the STPSC devices with different pairs of 1DPCs under a AM 1.5G illumination light source on the CIE 1931 color space. (b) The detailed section of CIE 1931 color space and (c) the CIE 1960 uniform color space under a AM 1.5G illumination light source.

Fig. 4a displays the CIE 1931 color space and represents the color coordinates position of each device illuminated by the AM 1.5G illumination light source. A detailed section of CIE 1931 color space is shown in Fig. 4b. When the light of the AM 1.5G illumination source transmits through the STPSC device without the 1DPCs (0 pair), the color of transmitted light is located on the CIE 1931 color space (0.3573, 0.3316), which is far away from the Planckian locus (black body locus).^30–33 For the CRI definition of CIE, the temperature of the nearest point of Planckian locus represents the CCT and the distance between device color coordinate and the CCT point on the Planckian locus is the DC. If the color coordinates are near to the Planckian locus, a higher DC and more accurate CRI will be obtained. Obviously, by incorporating the 1DPCs, the color coordinate was shifted towards the Planckian locus. Because of the uniform color of the CIE 1931 color space, the same distance on the CIE 1931 color space does not mean the same color change. Therefore, in order to illustrate the color change, the color coordinates (u, v) and the Planckian locus on the CIE 1960 uniform color space are given in Fig. 4c. It can be seen that the color coordinates of devices with 0, 2, 3, and 4 pairs of 1DPCs deviate from the Planckian locus, and the deviated distances are different from each other. For the device without the 1DPCs, the color coordinate is far away from the Planckian locus. However, for the devices with the 1DPCs, the color coordinates are located closer to the Planckian locus.

To testify the distance of the color coordinates and Planckian locus, the DC was calculated and is shown in Fig. 5a and Table 1. The CRI is only considered meaningful if the color coordinates lie within a chromaticity difference DC < 0.0054 from the Planckian locus on the CIE 1960 uniform color space, as recommended by the CIE. The DC of the device without the 1DPCs was about 0.0308, which is considerably higher than 0.0054. The variations in the DC are 0.0143, 0.0045 and 0.0050 corresponding to 2, 3, and 4 pairs of the 1DPCs. The smallest DC was obtained from the STPSC with 3 pairs of 1DPCs, which is the nearest to the Planckian locus. The CRI is an effective measurement of the ability for a light source. It can reveal the colors of various objects compared to an ideal or neutral light source, which is defined by the CIE. Here, the CRI accounts for the transmitted light color perception of the STPSC. STPSCs with a high CRI are desirable in window integration applications. The test sample method (TSM) was used to analyze the CRI perception.¹⁷ In CIE (1995) definition, the original fourteen test color samples (TCS) were taken from an early edition of the Munsell Atlas.²¹ First, we calculated the CRI_i of the first eight TCSs (1–8) and the CRI was the average of the CRI_i. The CRI_i of different devices illuminated using a AM 1.5G illumination light source are shown in Fig. 5b. It can be seen that the CRI totally presents an obvious enhancement when using 1DPCs.

Fig. 5 (a) The CRI and DC of different pairs of 1DPCs under illumination with a AM 1.5G illumination light source. The dashed horizontal line marks DC = 0.0054. (b) The CRI_i of TCS of different pairs of 1DPCs under illumination with a AM 1.5G illumination light source.

Table 1 The detailed performance parameters under 100 mW cm⁻² simulated AM1.5G light in ambient air and the CRI, DC of 0, 2, 3 and 4 pairs of 1DPCs devices

Devices	J sc (mA cm^−2)	V oc (V)	FF (%)	PCE (%)	CRI	DC
0 pair	9.47 ± 0.12	0.68 ± 0.01	53.3 ± 0.03	3.43 ± 0.11	77	0.0308
2 pairs	10.32 ± 0.10	0.68 ± 0.01	53.7 ± 0.02	3.77 ± 0.10	90	0.0143
3 pairs	10.88 ± 0.11	0.68 ± 0.01	54.2 ± 0.02	4.01 ± 0.11	91	0.0045
4 pairs	10.53 ± 0.10	0.68 ± 0.01	54.0 ± 0.03	3.87 ± 0.11	89	0.0050

---

Secondly, it was obvious that the CRI was improved using 1DPCs from 77 to 90, 91 and 89 corresponding to 2, 3, and 4 pairs of 1DPCs (Table 1). According to the DC definition, the CRI of 3 and 4 pairs of 1DPC devices were accurate. Hence, the optimal CRI of 91 with DC of 0.0045 was achieved using an STPSC with 3 pairs of 1DPCs. For the device with 4 pairs of 1DPCs, the transmittance spectrum of the STPSC was controlled by the photonic band gap and the amplitude of the transmittance spectrum increased.

Then, to understand the CRI calculation progress, TCSs color coordinates (x, y) on the CIE 1931 color space and (u, v) on the CIE 1960 uniform color space of the transmitted light with a AM 1.5G light source for the device are shown in Fig. 6, which is based on devices without and with 3 pairs of 1DPCs on the CIE 1931 and CIE 1960 color space. Along with the anti-clockwise starting point of the TCS circle, the one by one point represents TCS1–TCS8. According to the CRI calculation of the CIE definition, we use the BBR with the same CCT for reference. CCT is the color temperature of the nearest point on the Planckian locus, which most closely resembles a given stimulus under the specified viewing conditions and at the same brightness.¹⁹ Accordingly, we calculated the values of CCT and the results are 4050 K without the 1DPCs and 11 950 K with 3 pairs of 1DPCs. Therefore, according to the device CCT, different reference BBR sources were chosen and the color coordinates of the corresponding BBR reflected by each TCS were also calculated. We can visually determine the CRI by comparing the color coordinates distance of the same TCS of devices and BBR. It was evident that the TCSs color coordinates on both the CIE 1931 color space and CIE 1960 uniform color space of the device without 1DPCs are far away from the TCSs illuminated by the reference 4050 K BBR source in the two color spaces, which indicates the poor CRI of about 77. As implied by the excellent CRI of 91, the color coordinates of the TCSs illuminated by the transmitted light of device with 3 pairs of 1DPCs are close to the TCSs illuminated by the reference 11 950 K BBR both on the CIE 1931 color space and CIE 1960 uniform color space, clearly demonstrates the superior color perception of STPSC with 3 pairs of 1DPCs.

Fig. 6 (a) The TCSs color coordinates of the devices with 0 pairs and 3 pairs of 1DPCs under a AM 1.5G illumination light source and the TCSs color coordinates of the reference BBR on the CIE 1931 color space. (b) The CIE 1960 uniform color space. Along with the anti-clockwise starting point for the TCS circle, the one by one point represents TCS1–TCS8.

3.2 Device electrical properties

Fig. 7a and Table 1 show the J–V characteristics of the STPSCs with different pairs of 1DPCs and the error bars. The results were tested with more than 100 devices. Devices without 1DPCs exhibit a PCE of 3.43% with a short circuit current density (J_sc) of 9.47 mA cm⁻², open-circuit voltage (V_oc) of 0.68 V, and fill factor (FF) of 53.3%. Devices with 3 pairs of 1DPCs exhibit a PCE of 4.01% with J_sc of 10.88 mA cm⁻², V_oc of 0.68 V, and FF of 54.2%. It is clearly observed that the V_oc and FF of the devices are almost same, which can be attributed to the fact that FF and V_oc are independent of the 1DPC. Moreover, the J_sc is improved with the 1DPCs. In particular, for device with 3 pairs of 1DPC, a maximum J_sc enhancement is achieved.

Fig. 7 (a) J–V characteristics of devices, under 100 mW cm⁻² simulated AM 1.5G illumination light source in air. (b) IPCE characteristics of devices.

To certify the variation tendency of J_sc, the incident photo to current efficiency (IPCE) spectra of the devices were also measured and is shown in Fig. 7b. It can be seen that the IPCE values were totally enhanced in magnitude, and the variation tendency was in accordance with that found for J_sc. Furthermore, the devices with the 1DPCS exhibits an additional peak below 600 nm where there is a valley originally for the devices without the 1DPCs. Note that the positions and intensities of the emerging peaks are in good agreement with the reflection spectra of the 1DPCs. The IPCE enhancement can be attributed to the case that the strong reflection of 1DPCs results in a second-round absorption of the active layer in the wavelength range below 600 nm. With a combination of color investigation, it is demonstrated that the STPSCs structure with 1DPCs can achieve not only the improved CRI, but also an increased PCE.

4. Conclusions

In conclusion, we have achieved an improved CRI and enhanced PCE for low band gap STPSCs with 1DPCs. We designed the centre wavelength of the 1DPCs at 520 nm to level the concavo-convex transmittance spectrum induced by the absorption of PSBTBT:PC₆₀BM. By turning the pairs of 1DPCs, the optimal STPSC with 3 pairs of 1DPCs contributes an increased CRI from 77 to 91 under an AM 1.5G illumination light source, a suppressed DC from 0.0308 to 0.0045 and a enhanced PCE from 3.43% to 4.01% when compared to the device without the 1DPCs. It suggests the method of combining STPSCs with wavelength-dependent 1DPCs filters can achieve an almost perfect combination of optical, and electrical properties for their application requirements. We believe it will provide a potentially huge market in future BIPV applications.

Acknowledgements

The authors are grateful to the National Natural Science Foundation of China (Grant nos 61275035, 61370046, 51475200), Key Technology Research and Development Program of Changchun (no. 13KG66), Scientific Frontier and Interdiscipline Innovative Projects of Jilin University (Grant no. 2013ZY18) and Graduate Innovation Fund of Jilin University (Grant no. 2014021).

Notes and references

1. F. C. Krebs, Sol. Energy Mater. Sol. Cells, 2009, 93, 394.
2. Q. Yang, H. Song, B. Gao, Y. Wang, Y. Fu, J. Yang, Z. Xie and L. Wang, RSC Adv., 2014, 4, 12579.
3. H. Hsieh, Y. Sung, F. Hsu, K. Hsiao, Ya. Lee and Y. Chen, RSC Adv., 2015, 5, 1549.
4. J. Mescher, S. W. Kettlitz, N. Christ, M. F. Klein, A. Puetz, A. Mertens, A. Colsmann and U. Lemmer, Org. Electron., 2014, 15, 1476.
5. W. Zhang, M. Anaya, G. Lozano, M. E. Calvo, M. B. Johnston, H. Míguez and H. J. Snaith, Nano Lett., 2015, 15, 1698.
6. P. Romero-GÃģmez, F. Pastorelli, P. Mantilla-PÃĐrez, M. Mariano, A. MartÃnez-Otero, X. Elias, R. Betancur and J. Martorell, J. Photonics Energy, 2015, 5, 057212.
7. F. Guo, T. Ameri, K. Forberich and C. J. Brabec, Polym. Int., 2013, 62, 1408.
8. C.-C. Chen, L. Dou, R. Zhu, C.-H. Chung, T.-B. Song, Y. B. Zheng, S. Hawks, G. Li, P. S. Weiss and Y. Yang, ACS Nano, 2012, 6, 7185.
9. C. Y. Chang, L. Zuo, H. L. Yip, Y. Li, C. Z. Li, C. S. Hsu, Y. J. Cheng, H. Chen and A. K. Y. Jen, Adv. Funct. Mater., 2013, 23, 5084.
10. K.-T. Lee, J. Y. Lee, S. Seo and L. J. Guo, Light: Sci. Appl., 2014, 3, e215.
11. B. Park, S. H. Yun, C. Y. Cho, Y. C. Kim, J. C. Shin, H. G. Jeon, Y. H. Huh, I. Hwang, K. Y. Baik, Y. I. Lee, H. S. Uhm, G. S. Cho and E. H. Choi, Light: Sci. Appl., 2014, 3, e222.
12. W. Cao, J. D. Myers, Y. Zheng, W. T. Hammond, E. Wrzesniewski and J. Xue, Appl. Phys. Lett., 2011, 99, 023306.
13. W. Yu, L. Shen, P. Shen, Y. Long, H. Sun, W. Chen and S. Ruan, ACS Appl. Mater. Interfaces, 2014, 6, 599.
14. H. P. Kim, H. J. Lee, A. R. Bin Mohd Yusoff and J. Jang, Sol. Energy Mater. Sol. Cells, 2013, 108, 38.
15. F. Yan, J. Noble, J. Peltola, S. Wicks and S. Balasubramanian, Sol. Energy Mater. Sol. Cells, 2013, 114, 214.
16. T. Ameri, G. Dennler, C. Waldauf, H. Azimi, A. Seemann, K. Forberich, J. Hauch, M. Scharber, K. Hingerl and C. J. Brabec, Adv. Funct. Mater., 2010, 20, 1592.
17. A. Colsmann, A. Puetz, A. Bauer, J. Hanisch, E. Ahlswede and U. Lemmer, Adv. Energy Mater., 2011, 1, 599.
18. C.-C. Chen, L. Dou, J. Gao, W.-H. Chang, G. Li and Y. Yang, Energy Environ. Sci., 2013, 6, 2714.
19. J. Czolk, A. Puetz, D. Kutsarov, M. Reinhard, U. Lemmer and A. Colsmann, Adv. Energy Mater., 2013, 3, 386.
20. K.-S. Chen, J.-F. Salinas, H.-L. Yip, L. Huo, J. Hou and A. K. Y. Jen, Energy Environ. Sci., 2012, 5, 9551.
21. Commission Internationale de ĺÉclairage International 1995, ISBN 3-900-734-57-7.
22. ISO 10526:1999/CIE S005/E-1998, CIE Standard Illuminants for Colorimetry.
23. X. F. Li, J. D. Budai, F. Liu, J. Y. Howe, J. H. Zhang, X. J. Wang, Z. J. Gu, C. J. Sun, R. S. Meltzer and Z. W. Pan, Light: Sci. Appl., 2013, 2, e50.
24. J. Hou, H.-Y. Chen, S. Zhang, G. Li and Y. Yang, J. Am. Chem. Soc., 2008, 130, 16144.
25. L. Shen, G. Zhu, W. Guo, C. Tao, X. Zhang, C. Liu, W. Chen, S. Ruan and Z. Zhong, Appl. Phys. Lett., 2008, 92, 073307.
26. W. Yu, L. Shen, Y. Long, W. Guo, F. Meng, S. Ruan, X. Jia, H. Ma and W. Chen, Appl. Phys. Lett., 2012, 101, 153307.
27. J. Joannopoulos, P. Villeneuve and S. Fan, Nature, 1997, 386, 143.
28. J. Winn, Y. Fink, S. Fan and J. Joannopoulos, Opt. Lett., 1998, 23, 1573.
29. W. Yu, L. Shen, Y. Long, P. Shen, W. Guo, W. Chen and S. Ruan, Org. Electron., 2014, 15, 470.
30. C. Xiang, W. Koo, F. So, H. Sasabe and J. Kido, Light: Sci. Appl., 2013, 2, e74.
31. G. Wyszecki and W. S. Stiles, Color science: concepts and methods, quantitative data and formulae, Wiley-Interscience, 2nd edn, 2000, ISBN 0-471-39918-3.
32. M. Fröbel, T. Schwab, M. Kliem, S. Hofmann, K. Leo and M. C. Gather, Light: Sci. Appl., 2015, 4, e247.
33. J. H. Oh, S. J. Yang and Y. R. Do, Light: Sci. Appl., 2014, 3, e141.

---