Open Access
Add to BibSonomy
Abstract
Herein, we report the fabrication and characterization of a hierarchical TiO2 structure that exhibits reduced surface reflection. The hierarchical structure, which is a moth-eye-shaped array containing nanotubes, was fabricated by dry-etching a TiO2 nanotube layer, by using colloidal lithography. The fabricated structure shows a reduced reflectance, compared with that of non-patterned TiO2 nanotubes. This is because of the graded refractive index of the moth-eye pattern. Furthermore, we investigated the optical properties of gold-decorated moth-eye TiO2 nanotubes and found that the absorption, which was caused by the plasmonic resonance of gold nanostructures, was further enhanced by coupling with the light-trapping effect.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Titanium dioxide (TiO2) is one of the most studied and earth-abundant wide–band gap semiconductors for various applications such as photocatalysts, solar cells, photodetectors, gas sensors, UV blocking, and pigments [1–5]. Among the various wide–band gap semiconductors such as ZnO, GaN, and SiC, TiO2 exhibits advantages in terms of photo- and chemical stability, which are beneficial for long-term functioning [6,7]. However, large Fresnel reflections from the surface, due to the high refractive index (RI) of TiO2 (2.52 and 2.76 for the anatase and rutile phase, respectively), could decrease the performance of TiO2-based optical devices [8].
To reduce the surface reflection and improve the light-harvesting efficiency, various nanostructures have been exploited in previous studies [9,10]. Among them, graded-refractive-index (GRIN) structures, such as porous structure and nanocones, have a smooth RI gradient at the interface of two different media, resulting in suppressed reflections. One of the conventional fabrication methods for GRIN structures is nanoimprint lithography [11,12]. Desired patterns are transferred from molds onto a TiO2 nanoparticle–dispersed resin. Another strategy is to coat TiO2 nanoparticles on pre-patterned substrates. Zhang et al. deposited a monolayer of SiO2 nanospheres, as an antireflection (AR) coating, on glass substrates, and then coated smaller TiO2 particles on the surface of the SiO2 particles [8].
On the other hand, TiO2 nanotubes (NTs) fabricated by anodic anodization of Ti are known to be attractive structures for optoelectronic devices because of their facile fabrication, large surface area, and good electrical properties [13]. For example, when used as an electron transport layer for dye-sensitized or perovskite solar cells, TiO2 NT structures showed a higher charge collection efficiency compared with that of conventionally used nanoparticles [14,15]. To utilize the merits of TiO2 NT structures with enhanced light-harvesting efficiency, we fabricated moth-eye-patterned nanotubes (MPNTs) exhibiting a suppressed surface reflection compared with that of bare TiO2 NTs. By patterning the moth-eye-shaped GRIN structure on top of the NT layer, the reflection of the TiO2 NTs was reduced over the broadband spectrum in the UV to near-infrared range. After the growth of the NTs by anodic anodization, we etched the NT layer with a moth-eye-shaped structure by one-step reactive-ion etching (RIE) etching combined with colloidal lithography. The hierarchical structure, which is a nanocone array containing NTs at the interior, was fabricated simply. To the best of our knowledge, there have been no studies on direct patterning of NTs to reduce surface reflectance. Based on optical measurements, the average reflectance of the MPNTs in the 300–800 nm wavelength was 18.2%, which is a considerable reduction, compared with the 23.2% reflectance of NTs.
Furthermore, we exploited the morphologies and optical properties of two types of Au structures, Au nanoparticles (AuNP) and Au thin films, attached on the MPNT structure to study the combined effect of plasmonic resonance and the moth-eye structure. The AuNP/MPNT structure showed a lower reflection compared with that of the AuNP/NT structure because of suppressed backscattering as a result of the light trapping effect of the moth-eye structure. Moreover, when we coated the MPNTs with an Au thin film, the structure showed a broadband high absorptance of over 80% at 300–2500 nm.
2. Experimental
TiO2 NTs were prepared by anodic anodization of a Ti plate. The Ti (Grade 2) plate was electropolished in a mixed solution of 10 wt% perchloric acid (HClO4), 38 wt% ethylene glycol (C2H6O2), and methanol (CH3OH; 99.8%), at 20 V for 120 s with vigorous stirring [16]. After electropolishing, anodization at 40V was conducted in a solution of 0.5 wt% ammonium fluoride (NH4F) and 2 vol% H2O in ethylene glycol at 25 for 1 h. Then, the TiO2 layer was dried with an air-gun and completely peeled off using scotch tape. The adhesion between the TiO2 layer and Ti substrate was poor after anodization; therefore, the anodized TiO2 layer could be easily removed. Next, a second anodization was carried out under the same conditions as the first, for 15 min, resulting in a 5 µm thick TiO2 NT layer.
Then, we fabricated a moth-eye-shaped antireflective surface patterning on top of the TiO2 layer by RIE etching combined with colloidal lithography, as shown in Fig. 1 [17]. Firstly, we made an etch mask, which is a hexagonally closed packed monolayer of 500 nm polystyrene (PS) nanospheres, on top of the TiO2 NT layer. The solution of PS nanospheres (Thermo Fisher Scientific, 5050A) and ethanol (1:1 by volume) was spread on an air–water interface and transferred to the surface of the TiO2 layer [18]. Thereafter, we conducted RIE etching with a mixed plasma gas of SF6/O2/Ar (the flow rates of the gases were 50, 5, and 15 sccm, respectively). The working pressure and radio frequency (RF) power were 100 mTorr and 500 W, respectively.
3. Results and discussion
3.1 Moth-eye patterning by RIE etching
Figure 2 shows the SEM images of bare and surface-patterned NTs, obtained by scanning electron microscopy (SEM; JSM-7001F, JEOL Ltd.). In Fig. 2(a), the average diameter of the holes of bare NTs is ~50 nm. In Fig. 2(b), the PS nanospheres for the etch mask are very closely packed with a monolayer on top of the NT layer. Figure 2(c)–2(h) show the RIE-etched NTs for different etching times: 30, 60, and 90 s. During the etching, the nanosphere etch masks are scraped away and the transverse diameter of the nanospheres decreases gradually. As the diameter of the nanospheres decreases, the etched area of the TiO2 layer increases simultaneously. O2 gas reduces the size of the PS nanosphere etch mask, and SF6 gas effectively etches the TiO2 layer with the help of Ar gas, which maintains the plasma state. In our experiments, 60 s was the optimized etching time to fabricate the moth-eye pattern without any residual etch mask. Well-ordered periodic nanocone arrays were patterned on top of the NT layer, resulting in a hierarchical structure of nanocone arrays containing NTs. A shorter than 60 s etching time produces a frustum-like pattern, whereas a longer etching time destroys the moth-eye pattern. The aspect ratio of the nanocones was 2, and was measured from the cross-sectional SEM image obtained using a focused ion beam (FIB; Crossbeam 540, Zeiss) system, as shown in the inset of Fig. 2(f).
Fig. 2 SEM image of (a) bare NT, (b) close-packed nanospheres on NT, and NT after RIE etching for (c, d) 30 s, (e, f) 60 s, and (g, h) 90 s. (a, b, c, e, and g): top view; (d, f, and h): 30°- -tilted view. The inset in (f) shows the cross-sectional image of the MPNT layer. Each scale bar represents 500 nm.
3.2 Material characterization
The fabricated structure was annealed at 450 °C for 1 h to change the crystalline phase of TiO2 from amorphous to anatase. In our experiments, crystalline TiO2 showed a lower etch rate compared with that of amorphous TiO2; therefore, we annealed the sample after RIE etching. After annealing, we obtained X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) data to compare the crystallinity and elemental composition of the bare and RIE-etched NTs, as shown in Fig. 3. The XRD patterns in Fig. 3(a) show that NT and MPNT structures have dominant peaks that correspond to anatase TiO2. The 2θ values of the XRD peaks from the two structures were exactly similar. Furthermore, the XPS spectrum in Fig. 3(b) shows that NT and MPNT both have elemental compositions of C, O, and Ti. The C1s peak might have been generated because of carbon contamination in the XPS chamber or the TiO2 structure. Based on the XRD and XPS data, there were no notable changes in material properties as a result of RIE etching.
Fig. 3 Material properties of NT and MPNT. (a) XRD patterns (A: Anatase; R: Rutile; Ti: Titanium) and (b) XPS spectrum.
3.3 Antireflective effect of nano-porosity and moth-eye patterning
To evaluate the AR effect of MPNTs, we compared the RI of three TiO2 structures: bulk, NT, and MPNT. The RI of bulk TiO2 was taken from an earlier report and that of the porous NT structure was calculated using effective medium approximation (EMA) [19]. The porosity of the bare NTs, obtained from the binary SEM image in Fig. 4(a), was 0.25. The black parts are the holes, and the white part is TiO2. If we assume that a nanotube is an ideal two-dimensional structure, i.e., the inner diameter of the nanotube is constant along the whole nanotube, the volume ratio of air to TiO2 equals the area ratio of void to solid part, based on the SEM image. We used volume averaging theory (VAT) to calculate the effective RI of the NTs, because it was reported that VAT showed the best agreement for describing nanoporous TiO2, among several EMA methods such as the Maxwell–Garnett theory, Bruggeman EMA, and the parallel, series, Lorentz–Lorenz, and VAT models [20]. The real and imaginary parts of the effective RI ( and , respectively) were calculated using the following equations.
where where is the porosity, and are the real and imaginary parts of the RI of the continuous medium, and and are the real and imaginary parts of the RI of the dispersed medium. For the NT structure, the continuous medium is TiO2 and the dispersed medium is air. We assumed that the RI of air is similar to that of vacuum. Figure 4(b) shows the wavelength-dependent refractive index n and the extinction coefficient k of bulk TiO2 and NTs. Although the NTs have a lower RI value compared with that of the bulk, the RI of the NT structure in the UV spectrum is still above 2, which would generate more than 11% of Fresnel reflections at the air/TiO2 interface. Hence, it would be beneficial to reduce the surface reflection of the NTs using antireflective surface structures. Figure 4(c) shows the RI profiles at 350 nm for bulk TiO2, NTs, and MPNTs. The bulk and NT structures have a discrete RI at the interfaces, while MPNTs have a gradually changing RI between the air and TiO2 layer, which can reduce the Fresnel reflection. In addition to the GRIN effects, scattering or diffraction would occur when the wavelength is shorter than the lattice of the hexagonal array of MPNT. To suppress the scattering or diffraction for larger antireflection effect, less-ordered moth-eye patterns with smaller lattice would be beneficial [21,22].Fig. 4 (a) Binary SEM image of bare NT with a porosity of 0.25, (b) wavelength-dependent (350–800 nm) refractive index n and extinction coefficient k of bulk TiO2 and NT, and (c) refractive index profiles of bulk, NT, and MPNT structures at 350 nm wavelength in air.
For the further theoretical analysis, we performed 3D finite-difference time-domain (FDTD) simulations using a commercial software package (Lumerical FDTD Solution, Lumerical Inc.). We modelled the nanotube structures as hexagonal hole arrays with 70 nm diameter and 100 nm lattice, based on the SEM images. The maximum thickness of bulk, NT, and MPNT layer was 5 µm. The boundary conditions were set as periodic for the x- and y- directions and as perfectly matched layer (PML) conditions for the z-direction. The plane wave source in the range of 350 to 800 nm was incident to the structure with x-polarization. Figure 5(a) shows the reflectance spectrum of bulk, NT, and MPNT structures. Under the UV wavelength (λ<400 nm), the reflectance becomes lower in the order of bulk, NT, and MPNT. In the visible spectrum, Fabry-Perot fringes are shown in the reflectance spectra while the amplitude of the fringes is smaller for the NT and MPNT compared with the bulk structure. The simulated average reflectance of bulk, NT, and MPNT are 26%, 20%, and 14%, respectively, in 350 – 800 nm wavelength range. Also, we obtained FDTD-simulated electric field profiles |E|/|E0| (E is the amplitude of local electric field and E0 is that of incident field) at 350 nm wavelength. Figure 5(b) and Fig. 5(c) show the electric field profile of bulk and NT, respectively. For NT, more light goes into the TiO2 layer compared with the bulk. For MPNT, the electric field is enhanced in the moth-eye structure, as well as the light penetrates through the holes, as shown in Fig. 5(d). These FDTD simulation results show the beneficial effects of porous nanotube and moth-eye patterning for antireflection.
Fig. 5 (a) FDTD-simulated reflection spectra of bulk, NT, and MPNT structures on Ti substrate. E-field profiles of (b) bulk, (c) NT, and (d) MPNT, at 350 nm wavelength.
Following the theoretical calculations, we experimentally characterized the optical properties of the fabricated structures. The total reflection (sum of the specular and diffuse reflections) spectrum in the UV-Vis-NIR wavelength range was measured using a spectrophotometer (UV3600, Shimadzu Scientific Instruments) equipped with a BaSO4–coated integrating sphere (MPC–3100). Figure 6 shows the reflection spectrum of NTs and MPNTs. The moth-eye-patterned sample shows a lower reflection compared with that of the bare sample over the broadband spectrum of 300–800 nm. There were no Fabry-Perot fringes in the spectrum, which might be due to the defects in the TiO2 NT layer, thermal oxidation layer after annealing, and irregular surface of Ti substrate. The average reflectance of NTs and MPNTs are 23.2% and 18.2%, respectively, in 300–800 nm wavelength range. Thus, the average reduction in reflection is 21%. Under the UV wavelength of 350 nm, NTs show 14.1% reflectance, whereas MPNTs show 6.3%, which is a 55% reduction in reflection. In the UV regime, TiO2 layer absorbs the incident light. However, a considerable portion of the light is reflected from the surface of the unpatterned TiO2 layer because of the abrupt change of RI at air/TiO2 interface. By reducing the effective RI of the TiO2 layer, the RI change at the interface is reduced. As a result, reflected light decreases, and more light is penetrated and absorbed in the TiO2 layer. When the wavelength is larger than 550 nm, the incident light would not be absorbed from the TiO2 layer, as can be predicted from k value in Fig. 4(b). In that case, the reduced reflection, i.e. the absorption enhancement, would occur from the Ti substrate, because the Ti absorbs visible light.
To investigate how the size of moth-eye pattern affects the antireflection performance, we carried out FDTD simulation of MPNTs having three different period (P) of 300, 500, and 700 nm without changing the total height and aspect ratio of the moth-eye structure. Figure 7(a) shows the schematic illustration of the simulated structures. In Fig. 7(b), the reflectance in UV wavelength is almost zero for all three cases. Figure 7(c) to 7(e) exhibits the electric field profiles of MPNTs having three different P, under 350 nm illumination. When P increases, the location of the maximum enhancement of electric field moves downwards and the enhancement becomes larger. In the visible wavelength, several reflectance dips are exhibited when the P is 500 nm and 700 nm, as shown in Fig. 7(b). This phenomenon might be due to the absorption induced from the resonant modes of moth-eye structures [23]. It would be interesting to realize the simulated structures experimentally. For the fabrication, different sizes of the etch mask should be used for RIE etching, using optimized etching time.
Fig. 7 (a) Schematic illustration of MPTNs with three different period. (b) Simulated reflectance of MPNTs on Ti substrate. Electric field profiled of MPTNs with a period of (c) 300 nm, (d) 500 nm, (e) 700 nm, under 350 nm illumination.
3.4 Optical properties of Au-decorated structures
We exploited the morphology and optical properties of Au/MPNT because TiO2-based optoelectronic devices are often combined with Au plasmonic structures to enhance their efficiency or broaden the working wavelength [24,25]. Firstly, we fabricated Au nanoparticles (AuNP) by thermal annealing, on the MPNTs. Thin Au films were deposited on the TiO2 structure by sputtering, and they were annealed at 450 °C for 1 h. The nominal thickness of the sputtered Au films was 4 nm. Figure 8(a) shows the SEM image and photo of AuNP-decorated MPNTs (AuNP/MPNT). Through annealing, AuNPs with ~10 nm diameter were formed on the surface of the MPNTs. The color of the sample was dark red, which was induced by the localized surface plasmons of AuNPs. Figure 8(c) shows the total reflectance spectrum of four different structures: NTs, AuNP-decorated NTs (AuNP/NT), MPNTs, and AuNP-decorated MPNTs (AuNP/MPNT). Compared with that of NTs, AuNP/NT showed a lower reflectance, mainly in the visible-NIR wavelength, due to the plasmonic resonance of AuNPs. The center wavelength of the resonance is located at around 550 nm, although the bandwidth of resonance was very broad, because the size of the AuNPs was broadly distributed. Additionally, the absorption of the AuNPs due to interband transition reduces the reflectance in the UV-Vis-NIR regime. The AuNP/MPNT structure shows the lowest reflectance among the four structures, due to the combined effect of plasmonic resonance and the antireflection effect of the moth-eye structure. Using the moth-eye structure, the backscattering from the AuNPs and TiO2 layer can be suppressed, and thus, the reflections can be decreased compared with that for AuNP/NT [26]. The average reflectance of NTs, MPNTs, Au/NT, and Au/MPNT is 23.2%, 18.2%, 18.1%, and 12.5%, respectively. Secondly, we deposited 30 nm thick Au films on MPNTs without annealing (Au thin film/MPNT). Figure 8(b) shows the SEM image and photograph of the fabricated structure. The SEM image shows that the top surface of the MPNTs was completely covered with the Au film without the formation of nanoparticles or cracks. The fabricated structure looks very black. This color was the result of the broadband high absorption of Au thin film/MPNT, as shown in the reflectance/absorptance spectrum in Fig. 8(d). The structure had above 80% total absorptance in the UV-Vis-NIR range (300–2500 nm).
Fig. 8 SEM images and optical photographs of (a) AuNP/MPNT and (b) Au thin film/MPNT. Each scale bar indicates 200 nm. (c) Reflectance spectrum of NT and MPNT with AuNP decoration, and (d) reflectance/absorptance spectrum of Au thin film/MPNT.
4. Summary
In summary, we fabricated moth-eye-patterned TiO2 nanotubes having a graded refractive index, to suppress the large surface reflections from TiO2 films. The moth-eye structure was successfully fabricated by one-step RIE etching assisted by colloidal lithography, under optimized etching conditions. The optical properties of the fabricated structures were analyzed via theoretical studies and experimental measurements. Based on the measurements, the average reflectance of bare NTs, in the 300–800 nm range, was reduced by 21% due to the surface patterning. Furthermore, we coated the surface of MPNTs with the two different gold nanostructures (nanoparticles and thin films) to study the plasmonic effect of the moth-eye structure. As for the decoration with AuNPs, the AuNP/MPNT structure array showed a lower reflection than that of AuNP/NT, due to suppressed backscattering. After Au thin film coating, the MPNTs showed broadband high absorption characteristics over the UV-Vis-NIR wavelength range. We believe that the proposed moth-eye-patterned TiO2 nanotube structure exhibiting reduced surface reflection can be used as a platform for various TiO2-based applications with further experimentation.
Funding
National Research Foundation of Korea (NRF) (NRF-2013M3C1A3065045); Ministry of Science, ICT and Future Planning (CAMM-2014M3A6B3063712)
References
1. J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang, Y. Horiuchi, M. Anpo, and D. W. Bahnemann, “Understanding TiO2 photocatalysis: mechanisms and materials,” Chem. Rev. 114(19), 9919–9986 (2014). [CrossRef] [PubMed]
2. N. G. Park, G. Schlichthörl, J. van de Lagemaat, H. M. Cheong, A. Mascarenhas, and A. J. Frank, “Dye-sensitized TiO2 solar cells: structural and photoelectrochemical characterization of nanocrystalline electrodes formed from the hydrolysis of TiCl4,” J. Phys. Chem. B 103(17), 3308–3314 (1999). [CrossRef]
3. H. Xue, X. Kong, Z. Liu, C. Liu, J. Zhou, W. Chen, S. Ruan, and Q. Xu, “TiO2 based metal-semiconductor-metal ultraviolet photodetectors,” Appl. Phys. Lett. 90(20), 201118 (2007). [CrossRef]
4. H. Tang, K. Prasad, R. Sanjinés, and F. Lévy, “TiO2 anatase thin films as gas sensors,” Sens. Actuators B Chem. 26(1–3), 71–75 (1995). [CrossRef]
5. H. Yang, S. Zhu, and N. Pan, “Studying the mechanisms of titanium dioxide as ultraviolet-blocking additive for films and fabrics by an improved scheme,” J. Appl. Polym. Sci. 92(5), 3201–3210 (2004). [CrossRef]
6. K. Sridharan, E. Jang, Y. M. Park, and T. J. Park, “Superior photostability and photocatalytic activity of ZnO nanoparticles coated with ultrathin TiO2 layers through atomic-layer deposition,” Chemistry 21(52), 19136–19141 (2015). [CrossRef] [PubMed]
7. A. G. Scheuermann and P. C. McIntyre, “Atomic Layer Deposited Corrosion Protection: A Path to Stable and Efficient Photoelectrochemical Cells,” J. Phys. Chem. Lett. 7(14), 2867–2878 (2016). [CrossRef] [PubMed]
8. X.-T. Zhang, O. Sato, M. Taguchi, Y. Einaga, T. Murakami, and A. Fujishima, “Self-Cleaning Particle Coating with Antireflection Properties,” Chem. Mater. 17(3), 696–700 (2005). [CrossRef]
9. W. Zhu, F. Xiao, I. D. Rukhlenko, J. Geng, X. Liang, M. Premaratne, and R. Jin, “Wideband visible-light absorption in an ultrathin silicon nanostructure,” Opt. Express 25(5), 5781–5786 (2017). [CrossRef] [PubMed]
10. Q. Chen, J. Gu, P. Liu, J. Xie, J. Wang, Y. Liu, and W. Zhu, “Nanowire-based ultra-wideband absorber for visible and ultraviolet light,” Opt. Laser Technol. 105(8), 102–105 (2018). [CrossRef]
11. S. M. Kang, S. Jang, J.-K. Lee, J. Yoon, D.-E. Yoo, J.-W. Lee, M. Choi, and N.-G. Park, “Moth-eye TiO2 layer for improving light harvesting efficiency in perovskite solar cells,” Small 12(18), 2443–2449 (2016). [CrossRef] [PubMed]
12. S. Jang, S. M. Kang, and M. Choi, “Multifunctional Moth-Eye TiO2/PDMS Pads with High Transmittance and UV Filtering,” ACS Appl. Mater. Interfaces 9(50), 44038–44044 (2017). [CrossRef] [PubMed]
13. P. Roy, S. Berger, and P. Schmuki, “TiO2 nanotubes: synthesis and applications,” Angew. Chem. Int. Ed. Engl. 50(13), 2904–2939 (2011). [CrossRef] [PubMed]
14. G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese, and C. A. Grimes, “Use of highly-ordered TiO2 nanotube arrays in dye-sensitized solar cells,” Nano Lett. 6(2), 215–218 (2006). [CrossRef] [PubMed]
15. X. Gao, J. Li, S. Gollon, M. Qiu, D. Guan, X. Guo, J. Chen, and C. Yuan, “A TiO2 nanotube network electron transport layer for high efficiency perovskite solar cells,” Phys. Chem. Chem. Phys. 19(7), 4956–4961 (2017). [CrossRef] [PubMed]
16. N. S. Peighambardoust and F. Nasirpouri, “Electropolishing behaviour of pure titanium in perchloric acid–methanol–ethylene glycol mixed solution,” Trans. IMF 92(3), 132–139 (2014). [CrossRef]
17. H. Park, D. Shin, G. Kang, S. Baek, K. Kim, and W. J. Padilla, “Broadband optical antireflection enhancement by integrating antireflective nanoislands with silicon nanoconical-frustum arrays,” Adv. Mater. 23(48), 5796–5800 (2011). [CrossRef] [PubMed]
18. M. Retsch, Z. Zhou, S. Rivera, M. Kappl, X. S. Zhao, U. Jonas, and Q. Li, “Fabrication of large-area, transferable colloidal monolayers utilizing self-assembly at the air/water interface,” Macromol. Chem. Phys. 210(3–4), 230–241 (2009). [CrossRef]
19. M. Mosaddeq-ur-Rahman, G. Yu, T. Soga, T. Jimbo, H. Ebisu, and M. Umeno, “Refractive index and degree of inhomogeneity of nanocrystalline TiO[sub 2] thin films: Effects of substrate and annealing temperature,” J. Appl. Phys. 88(8), 4634 (2000). [CrossRef]
20. A. Garahan, L. Pilon, J. Yin, and I. Saxena, “Effective optical properties of absorbing nanoporous and nanocomposite thin films,” J. Appl. Phys. 101(1), 014320 (2007). [CrossRef]
21. P. I. Stavroulakis, S. A. Boden, T. Johnson, and D. M. Bagnall, “Suppression of backscattered diffraction from sub-wavelength ‘moth-eye’ arrays,” Opt. Express 21(1), 1–11 (2013). [CrossRef] [PubMed]
22. L. W. Chan, D. E. Morse, and M. J. Gordon, “Moth eye-inspired anti-reflective surfaces for improved IR optical systems & visible LEDs fabricated with colloidal lithography and etching,” Bioinspir. Biomim. 13(4), 041001 (2018). [CrossRef] [PubMed]
23. G. Kang, H. Park, D. Shin, S. Baek, M. Choi, D. H. Yu, K. Kim, and W. J. Padilla, “Broadband light-trapping enhancement in an ultrathin film a-Si absorber using whispering gallery modes and guided wave modes with dielectric surface-textured structures,” Adv. Mater. 25(18), 2617–2623 (2013). [CrossRef] [PubMed]
24. Z. Zhang, L. Zhang, M. N. Hedhili, H. Zhang, and P. Wang, “Plasmonic gold nanocrystals coupled with photonic crystal seamlessly on TiO2 nanotube photoelectrodes for efficient visible light photoelectrochemical water splitting,” Nano Lett. 13(1), 14–20 (2013). [CrossRef] [PubMed]
25. A. Furube and S. Hashimoto, “Insight into plasmonic hot-electron transfer and plasmon molecular drive: new dimensions in energy conversion and nanofabrication,” NPG Asia Mater. 9(12), e454 (2017). [CrossRef]
26. C. L. Tan, S. J. Jang, and Y. T. Lee, “Localized surface plasmon resonance with broadband ultralow reflectivity from metal nanoparticles on glass and silicon subwavelength structures,” Opt. Express 20(16), 17448–17455 (2012). [CrossRef] [PubMed]
