All-Dielectric Metasurfaces Based on Cross-Shaped Resonators for Color Pixels with Extended Gamut
Vishal Vashistha, Gayatri Vaidya, Ravi S Hegde, Andriy E Serebryannikov, Nicolas Bonod, Maciej Krawczyk
AAll-Dielectric Metasurfaces Based onCross-Shaped Resonators for Color Pixels withExtended Gamut
Vishal Vashistha, ∗ , † , (cid:107) Gayatri Vaidya, ‡ , (cid:107) Ravi S. Hegde, ¶ Andriy ESerebryannikov, † Nicolas Bonod, § and Maciej Krawczyk ∗ , † † Faculty of Physics, Adam Mickiewicz University in Poznan, Poland ‡ Centre of Excellence in Nanoelectronics - CEN, IIT Bombay, India 400076 ¶ Indian Institute of Technology, Gandhinagar, India 382355 § Aix Marseille Univ, CNRS, Centrale Marseille, Institut Fresnel, 13013 Marseille, France (cid:107) these authors contributed equally to this work.
E-mail: [email protected]; [email protected]
Abstract
Printing technology based on plasmonic structures has many advantages over pig-ment based color printing such as high resolution, ultra-compact size and low powerconsumption. However, due to high losses and broad resonance behavior of metals inthe visible spectrum, it becomes challenging to produce well-defined colors. Here, weinvestigate cross-shaped dielectric nanoresonators which enable high quality resonancein the visible spectral regime and, hence, high quality colors. We numerically predictand experimentally demonstrate that the proposed all-dielectric nanostructures exhibithigh quality colors with selective wavelengths, in particular, due to lower losses as com-pared to metal based plasmonic filters. This results in fundamental colors (RGB) with a r X i v : . [ phy s i c s . op ti c s ] F e b igh hue and saturation. We further show that a large gamut of colors can be achievedby selecting the appropriate length and width of individual Si nanoantennas. More-over, the proposed all-dielectric metasurface based color filters can be integrated withthe well matured fabrication technology of electronic devices. Keywords
All-dielectric nanophotonics, color filter, plasmonics, structural colors, nanoantenna.With tremendous changes in nanotechnology over past few decades, it becomes possibleto fabricate devices which promise to revolutionize many areas. Examples include ultra-thinplanar lens, optical sensing, photo-voltaic devices, non-fading colors, and variousholography based devices. In particular, color pixels using nanoparticles have gained sig-nificant attention in recent years because of several advantages over pigment based colorprinting techniques like high resolution, high contrast, everlasting colors, significant lowpower consumption, and recyclability of product. The concept of structural color printingis inspired by observations in nature, such as morpho butterflies, beetles, and the feathersof peacocks.
However, these colors are highly sensitive to the variations in the angle ofincidence, shape, and size of the nanostructure. To make this plasmonics based structuraltechnology more mature, its angle dependency, sensitivity to polarization, and ease offabrication must be taken into account. In recent years, many efforts have been done to studythe aforementioned issue in plasmonic color printing.
Earlier, the most commonly usedmaterials for plasmonic nanostructure based pixels have been gold and silver.
Gold hasinterband transition in the lower visible regime, while silver is suitable for the entire visiblerange but is susceptible with the native oxide that spoils the stability of colors. Moreover,gold and silver are not economical for large scale integration. Aluminum is probably themost prominent candidate. It is more robust and economical for large-scale fabrication. However, it shows lower quality ( i.e. , broader) resonance in the visible spectrum than gold orsilver, especially at 800 nm wavelength, where interband transition takes place. Ultimately,2ll these metal based plasmonic devices show significant losses within the visible spectrum.On the other hand, all-dielectric metasurfaces can be a promising solution with sig-nificant advantages over metallic nanostructures such as high quality resonances and lowintrinsic ohmic losses. Silicon based all-dielectric devices have been reported for localmanipulation by wavefronts, such as beam diversion, vortex plates and light focusing usingmeta-lenses.
The advantages of Si nanodisks are high refractive index and ease offabrication with well established CMOS technology. Interestingly, the high refractive indexallows to manipulate by magnetic and electric components of light simultaneously. In the caseof metal based nanoantennas, absorption losses can be significant at visible spectrum, whileinteraction with magnetic component of the incident beam requires more complex shapes.Recently, an investigation has been conducted to demonstrate the possibility of using silicon-aluminum hybrid nanodisks to create colors of high quality. Silicon nanoparticles wereproposed as a valuable alternative to plasmonic nanoantennae for the design of color pix-els. However, the potential of all-dielectric resonance structures is presently very farfrom being fully estimated and exploited.In this work, we propose a systematic approach to build color filters by using advantagesof cross-shaped Si nanoresonators, which are closely spaced to each other to create a meta-surface. Recently reported numerical studies of the nanocross geometry have indicatedthat a broader gamut of colors is possible in comparison to simpler shapes like the cylinder(disk). The main goal is to obtain a high quality (narrow) resonance throughout the visiblespectrum that enables an extended gamut with colors of high purity. It is known that Si nanostructures of different shapes typically offer an opportunity to excite individual electrictype and magnetic type Mie resonances, or both resonances simultaneously. In fact, it hasbeen demonstrated that by tuning the aspect ratio carefully, one can overlap both resonancesto achieve near unity transmission. In this paper, the all-dielectric metasurfaces are used3n reflection mode. A very confined energy is concentrated within the structure due to thehigh quality of the used Mie resonances.The main hypothesis that we follow here is based on the expectation that a proper manip-ulation by the selected Mie resonances may enable desired improvements of the resulting res-onance quality owing to better confinement of resonance fields and, simultaneously, removalof secondary (unwanted) spectral features, so that enrichment of colors can be achieved. Wedecided in favor of cross-shaped Si nanoresonators as building elements, which are expectedto be suitable for achievement of the goals of this study. Each of them is made of two iden-tical orthogonal rectangle-shaped Si nanoantennas. In this case, resonances are governedby cross-shaped nanoantennas and thus, colors can be controlled via all three geometricalparameters of individual nanoantennas. This gives a new degree of freedom as comparedto the nanodisks, that is highly demanded for efficient optimization. Using the suggestedapproach, we predict by simulations and confirm experimentally that one can easily achievea high quality resonance for the entire visible spectrum by carefully choosing the length andwidth of the cross-shaped nanoresonators. Results
Let us start from the general geometry and basic operation principles of the proposed devices.Figure 1(a) presents the perspective view of the proposed all-dielectric metasurface togetherwith some details of geometry. The cross-shaped Si nanoresonators are deposited on top ofthe quartz substrate (see Methods of fabrication). The height of nanoantennae is selected as140 nm (in subwavelength range). Figure 1(b) represents the top view of SEM image of thedevice. A 45 ◦ cross section view is also added in the inset for the same fabricated device.For the studied Si structure, extinction cross section spectrum is presented in Fig. 1(c).Two resonance peaks are observed at 465 nm and 520 nm . They can be tuned throughout4 s u b s t r a t e t h W L Si x y z (a) (b)(c)
250 nm (d)
400 450 500 550 600 650 7000.00.20.40.60.81.0
400 450 500 550 600 650 7000.00.20.40.60.81.0 E x t i n c t i on c r o ss s e c t i on ( a . u . ) Wavelength (nm)
ECS (Al)ECS (Si) R e f l ec t a n ce ( a . u . ) Wavelength (nm)
AlSi
Figure 1:
Perspective view and SEM images of the all-dielectric metasurface, ex-tinction cross section (ECS) spectra and reflectance spectra, and reflectance vs polarization angle. (a) Schematic representation of the array of cross-shaped Si nanores-onators on top of the quartz substrate. The thickness of the substrate t = 275 µ m . Foreach nanoantenna, height h = 140 nm , length L and width W are scaled to achieve differentcolors. The center-to-center distance between the two nanoresonators (lattice constant) is P = 250nm . (b) Top view of SEM images of the fabricated structure with P = 250 nm . A45 ◦ cross section view is added in the inset. (c) ECS spectra in case of Si and Al nanoan-tennae. Two peaks arising in the former case are due to electric type and magnetic typeresonance (see supplementary information for the field patterns), while there is only singlebroad resonance in the latter case. The inset shows the reflectance spectra for the same twostructures. The length, width and height are 100 nm , 50 nm and 140 nm , respectively. (d) Acolormap of simulated reflectance spectra of the Si based metasurface at polarization anglevaried from 0 ◦ to 360 ◦ .the visible range by changing the length-to-width aspect ratio of individual rectangle-shapednanoantennas. The Si nanoresonator dimensions have been optimized to excite these tworesonances as close as possible but without a full overlapping. In addition, the criterium5f minimizing unwanted spectral features has been applied in order to obtain more gradualbehavior in the working spectral range. As follows from the obtained simulation results,optimization yields a resonance range that is narrower and, thus, corresponds to a resonanceof higher quality, as compared to the case of Al cross-shaped nanoantennae, see Fig. 1(c). Wehave also compared the simulated reflectance spectra for the metal and Si based structures atthe same dimensions [see Fig. 1(c), inset]. These results confirm that the metal nanostructurefeatures broader resonances than the engineered Si one. An important advantage of cross-shaped nanoantennae is that they preserve the polarization independence. As an example,Fig. 1(d) presents the simulated reflectance spectrum for the entire range of polarizationangle variation and entire wavelength range considered. The obtained results confirm thatthere is no change in the reflectance spectrum when the polarization angle is varied.Since a specific color results from resonant interaction of light with nanoresonators, itcan be obtained from adjustment of geometrical parameters that properly affect spectrallocations and properties of Mie resonances. The possibility of obtaining multiple colors withthe aid of metasurfaces like that in Fig. 1(a) and (b) is demonstrated in Fig. 2. The lengthand width of rectangle-shaped Si nanoantennae are simultaneously linearly scaled in orderto tune the electric and magnetic type resonances in the entire visible spectrum from 400 nm to 700 nm , as shown in Fig. 2(a) for P = 250nm . A commercial-grade simulator based onthe finite-difference time-domain method is used to perform the calculations. They areconducted for a unit cell with periodic boundary conditions, and varied lattice constantfrom 250 nm to 350 nm , by keeping the periodicity in the subwavelength range (see Methods,Simulation). Each spectral zone in Fig. 2(a) corresponds to a specific color. It is clearlyseen that the electric and magnetic type resonances can be tuned through the entire visiblewavelength spectrum, as desired. Conversion of reflectance spectra into colors on CIE1931chromaticity diagram can be performed, in the general case, by using an open source Pythonprogram. The results of conversion of the spectra shown in Fig. 2(a) are presented inFig. 2(b). Generally, a higher quality of resonances corresponds to a better approaching6o the boundaries of the chromaticity diagram and, hence, enable higher quality and widergamut of colors. Complete details about color visualization using reflectance spectra are givenin supplementary information under section color representation from reflectance spectra.By operating the metasurface in reflection mode, a broad spectrum of colors for highlyselective wavelengths ( i.e. , high quality colors) can be obtained. In principle, colors canbe generated by using either additive or subtractive approach. Here, we have used theadditive approach. Ideally, the reflection spectrum must be as narrow as possible in order togenerate a very specific color. A narrower resonance represents a more specific wavelengthcolor, whereas the amplitude of the peak decides the saturation level of the color. Withthe aid of high quality narrow resonances, we improve the approaching to the boundariesof CIE-1931 chromaticity diagram, so a color of higher quality and a wider gamut of colorscan be obtained, as desired. We experimentally found that different colors can be obtainedat different values of period ( P , lattice constant), which correspond to the scaled length ( L ) and width ( W ) of the nanoantenna, see Fig. 2(c). Each square in Fig. 2(c) correspondsto a unique set of geometrical parameters. The lowest series of the squares shown herecorresponds to the structures, for which reflectance spectra are presented in Fig. 2(a). Thus,the resonance region corresponds to different colors at different values of P , see Fig. S4in supplementary information. This dependence occurs owing to the coupling of resonancefields of nanoresonators. The use of larger values of P allows us to create a richer variety ofcolors, as we have more choices to increase the length and width. We have observed differentcolors under optical microscope due to variations in lattice constant ( P ) from 250 nm to350 nm , see Fig. 2(c). The lattice constant was increased here by a reasonable increment of20 nm to make it feasible for fabrication process. Although it might be hard to distinguishbetween the highly saturated colors in Fig. 2(c), the reflectance spectra in Fig. 2(a) and thecorresponding CIE-1931 chromaticity diagram in Fig. 2(b) give us a clear picture about it.In fact, a color gamut can be possible by making a matrix between the scaled lengths andwidths. 7he fact that two resonances, which are observed in Fig. 2(a) at different values of L and W , are closely spaced makes fabrication of a particular color possible, that is unlikely incase of metal based plasmonic structures, because they show a broad resonance. Moreover,it is possible to create a selective wavelength color due to sharp resonances, particularly inthe lower part of the visible spectrum. It is observed in Fig. 2(a) that as we increase the sizeof cross-shaped resonators some additional Mie resonances are also excited, in coincidencewith the predictions based on the simulation results. These resonances reduce the hue andsaturation of red color, because of mixing contribution of different frequencies. So the redcolor seems to be the most difficult one to fabricate. Below, we will show that in spite ofthe above-mentioned difficulties the suggested structure allows us creating fairly red colorsby carefully adjusting the values of P , L , and W . Thanks to this adjustment, the unwantedeffect of higher-order resonances can be minimized.The three primary colors (RGB) represent the fundamental unit for color printing tech-nology. All the other colors in the RGB gamut can be derived by mixing the primary colorsappropriately. Figure 3 presents the results of a detailed experimental demonstration ofthe suggested devices in the form of pixels. A dual characterization is done to ensure theresults by measuring the reflectance spectra of the samples with the aid of a home-madecustomized setup and observing the colors directly under optical microscope (see Methods,Optical characterization). Figure 3(a) shows the experimental and simulated reflectancespectra for highly saturated primary colors. These results show good agreement with eachother. Figure 3(b) shows the SEM images obtained at different sizes of nanoantennas. In-sets are added to the SEM images to show the corresponding colors visible under opticalmicroscope, which are associated with the different sizes of the cross-shaped nanoresonators.Finally, these three primary colors are fabricated in a form of pixel, being the main com-ponent of any display device. The size of each square block is 50 µ m . Details of the usedfabrication method are given at the end of the paper. The optical microscope images shown8 caling factor (k) P e r i o d i c i t y ( - n m ) C I E y CIE 1931 Chromaticity Diagram - CIE 1931 2 ◦ Standard Observer (b)(c)
400 450 500 550 600 650 700
Wavelength (nm) R e f l e c t an c e ( a . u . ) Periodicity (250 nm) (a) ^ ^ Figure 2:
Reflectance spectra (simulation), corresponding chromaticity diagram,and photograph of experimental images of the array visible under optical mi-croscope. (a) Unit-cell simulation results for cross shaped Si nanoresonators on quartzsubstrate; the initial values of geometrical parameters are L = 65 nm , W = 35 nm , and P =250 nm . L and W are linearly scaled from 65 nm to 195 nm and 35 nm to 105 nm , respectively,from bottom to top. The resonances are redshifted in the visible regime, as schematicallyshown by arrows. (b) Representation of reflectance spectra on standard CIE 1931 chro-maticity diagram for P = 250 nm . (c) Experimental colors visible under optical microscopefor different values of P which are varied from 250 nm (the lowest series) to 350 nm (themost upper series) with step of 20 nm (from the lowest series to the most upper one); L and W are linearly scaled from 65 nm to 260 nm and 35 nm to 140 nm , respectively; with k = L/ min( L ) = W/ min( W ) .
9n Fig. 3(c) confirm the quality of highly saturated primary colors, which is an importantadvantage of the suggested all-dielectric metasurface based pixels over the existing plasmon-ics devices. A CIE 1931 chart is used to represent the simulated and experimental spectraof the primary colors, see Fig. 3(d). One can see a very small shift in color spectrum, whichmight come from fabrication imperfections. It is noticeable that there is good coincidencebetween two sets of experimental results.
Conclusion
Polarization insensitive all-dielectric metasurfaces based on 2D arrays of cross-shaped Si nanoresonators have been proposed to realize color filters with extended gamut for the en-tire visible spectrum. A numerical investigation has been carried out that demonstrates theprincipal possibility of obtaining high-purity colors by means of optimization of resonanceproperties, which can be realized by a relatively simple adjustment of the structural param-eters. The role of existence and properties of the dual resonance, which is achieved at apartial overlapping of electric type and magnetic type resonances, and that of suppression ofunwanted spectral features in the obtaining of these advancements have been clarified. Theutilized resonances can be tuned by changing the length-to-width aspect ratio of individualrectangle-shaped nanoantennas. This concept has been used to design and fabricate the colorfilters. Our simulation results reasonably agree with the experimental ones. Some differencesshould be noticed that may be connected with fabrication complexity of the structure. Theexperimentally demonstrated possibility of obtaining high quality (narrow) resonances, whichenable high quality colors, is the most important result of this work. We have demonstratedthe wide variety of colors for different periodicity and size of cross-shaped nanoresonators.Additionally, we have demonstrated the primary colors painting in the form of pixels. Thesecolors show high saturation and hue value. In fact, our device is capable to produce a largepanel color in the visible regime with strong spectral selectivity, provided that the nanoan-10
00 450 500 550 600 650 7000.00.20.40.60.81.0 R e f l ec t a n ce ( a . u . ) Wavelength (nm)
SimulationExperimental
400 450 500 550 600 650 7000.00.20.40.60.81.0 R e f l ec t a n ce ( a . u . ) Wavelength (nm)
SimulationExperimental
400 450 500 550 600 650 7000.00.20.40.60.81.0 R e f l ec t a n ce ( a . u . ) Wavelength (nm)
ExperimentalSimulation
250 nm 250 nm 250 nm
20 micron (a)(c) (d) (iii)(ii)(i) (b) (iii)(i) (ii)(i) (ii)
Figure 3:
Simulation and experimental results for primary colors with SEM im-ages. (a) Simulated and experimental reflectance spectra. The dimensions of cross-shaped Si nanoresonators (i) L = 85 nm and W = 46 nm for blue color. (ii) L = 114 nm and W = 62 nm for green color and (iii) L = 215 nm and W = 116 nm for red color. The latticeconstant is 250 nm for all three cases. (b) SEM images of the fabricated structures with theinset view of associated colors. (c) A photograph taken from Nikon camera attached with50 × lens with N A = 0 . and the enlarged SEM image of the nanostructure. (d) A CIE1931 chromaticity diagram that is used to visualize the simulated and measured colors.tenna aspect ratio is properly chosen. By carefully controlling the balance between desiredand unwanted Mie resonances, one can further optimize the color filter, especially in darkzone of red color. Since the a − Si (amorphous- Si ) is the most suitable material for large-scale fabrication with the existing technology, it can potentially be used for making low-cost,eco-friendly, high quality, long lasting painting possible for mass production in the future.11 ethods Simulations.
We have used Lumerical FDTD solver to study metasurfaces comprisingthe cross-shaped nanoresonators on a dielectric substrate. The materials used for substrateand cross-shaped nanoresonators are SiO and Si , respectively. The material parametersare taken from default the material library of the used software. A plane wave ranging from400 nm to 700 nm is illuminated from the top of the structure. Periodic boundary conditionsare used in the unit cell along x and y directions. Perfect matching layer (PML) boundaryconditions were used in the z directions to avoid any reflection. The reflectance spectraare simulated by considering a unit cell (one cross-shaped nanoresonator on substrate) withperiodic boundary conditions in x and y directions. Device fabrication.
A piranha cleaned quartz sample (275 µ m thick) is used to fabricatethe device. We have deposited a thin layer of 140 nm amorphous Si using ICPCVD tool at300 ◦ Celsius with 150W added microwave power. A single-layer PMMA photoresist is usedfor patterning cross-shaped nanoresonators by using Raith 150-Two EBL tool. An electronicmask is designed using an open source Python program. The exposed sample is developedusing MIBK-IPA (1:3) and an IPA solution for 45 s and 15 s , respectively. A thin layer ofmetal (5 nm Cr as adhesion layer and 40 nm Au ) is deposited to transfer the pattern onmetal layer for lift-off process using four target evaporators. After lift-off, the sample isetched using plasma asher to get the final pattern. A process flow chart with step by stepdetails is available in supplementary information. Optical characterization.
A dual optical characterization is done to ensure the results.The sample is placed under Olympus optical microscope and illuminated with white lightwithout filter. The colors can be directly seen under optical microscope. The reflectancespectra are measured using a home-made customized setup. A HL 2000 halogen lamp sourceis coupled with an optical fiber to illuminate the sample in the visible range, i.e. , from 400 nm to 700 nm . A 50 × objective lens with N A = 0 . is used to get tight focusing of light on thesample. The reflectance spectra are measured using the same objective lens. All the collected12ata are normalized with respect to the bare quartz sample. A Nikon camera attached withassembly is used to take the photograph of the illuminated area. Acknowledgement
RSH acknowledges support from Department of Science and Technology, India under theExtramural Research Grant no. SB/S3/EECE/0200/2015. RSH, VV and GV acknowledgesupport from Indian Nanoelectronics Users Program under grant nos. P643987963 andP875860276. The work was partially supported by the National Science Centre Poland forOPUS grant No. 2015/17/B/ST3/00118(Metasel) and by the European Union Horizon2020research and innovation program under the Marie Sklodowska-Curie grant agreement No.644348 (MagIC). Authors thank all the members of CEN laboratory, IIT Bombay whohelped us directly or indirectly while doing nanofabrication work. Special thanks to Dr. KNageshwari and Dr. Ritu Rashmi for providing necessary facilities and regular advice.
Supporting Information Available
The following files are available free of charge.
References (1) Kildishev, A. V.; Boltasseva, A.; Shalaev, V. M. Planar photonics with metasurfaces.
Science , , 1232009.(2) Yu, N.; Capasso, F. Flat optics with designer metasurfaces. Nat. Materials , ,139–150.(3) Adato, R.; Yanik, A. A.; Amsden, J. J.; Kaplan, D. L.; Omenetto, F. G.; Hong, M. K.;13rramilli, S.; Altug, H. Ultra-sensitive vibrational spectroscopy of protein monolayerswith plasmonic nanoantenna arrays. Proc. Nat. Acad. Sci. , , 19227–19232.(4) Liu, N.; Mesch, M.; Weiss, T.; Hentschel, M.; Giessen, H. Infrared perfect absorber andits application as plasmonic sensor. Nano Lett. , , 2342–2348.(5) Atwater, H. A.; Polman, A. Plasmonics for improved photovoltaic devices. Nat. Mate-rials , , 205–213.(6) Chen, X.; Jia, B.; Saha, J. K.; Cai, B.; Stokes, N.; Qiao, Q.; Wang, Y.; Shi, Z.; Gu, M.Broadband enhancement in thin-film amorphous silicon solar cells enabled by nucleatedsilver nanoparticles. Nano Lett. , , 2187–2192.(7) James, T. D.; Mulvaney, P.; Roberts, A. The plasmonic pixel: large area, wide gamutcolor reproduction using aluminum nanostructures. Nano Lett. , , 3817–3823.(8) Almeida, E.; Bitton, O.; Prior, Y. Nonlinear metamaterials for holography. Nat. Com-mun. , , 12533.(9) Ni, X.; Kildishev, A. V.; Shalaev, V. M. Metasurface holograms for visible light. Nat.Commun. , , 3807.(10) Tan, S. J.; Zhang, L.; Zhu, D.; Goh, X. M.; Wang, Y. M.; Kumar, K.; Qiu, C.-W.;Yang, J. K. Plasmonic color palettes for photorealistic printing with aluminum nanos-tructures. Nano Lett. , , 4023–4029.(11) Clausen, J. S.; Højlund-Nielsen, E.; Christiansen, A. B.; Yazdi, S.; Grajower, M.;Taha, H.; Levy, U.; Kristensen, A.; Mortensen, N. A. Plasmonic metasurfaces for col-oration of plastic consumer products. Nano Lett. , , 4499–4504.(12) Srinivasarao, M. Nano-optics in the biological world: beetles, butterflies, birds, andmoths. Chemical Rev. , , 1935–1962.1413) Vukusic, P.; Sambles, J.; Lawrence, C.; Wootton, R. Structural colour: now you see it—now you don’t. Nature , , 36–36.(14) Kinoshita, S.; Yoshioka, S.; Miyazaki, J. Physics of structural colors. Repor. Prog. Phys. , , 076401.(15) Gralak, B.; Tayeb, G.; Enoch, S. Morpho butterflies wings color modeled with lamellargrating theory. Opt. Express , , 567–578.(16) Wu, Y.-K. R.; Hollowell, A. E.; Zhang, C.; Guo, L. J. Angle-insensitive structuralcolours based on metallic nanocavities and coloured pixels beyond the diffraction limit. Sci. Rep. , , 1194.(17) Højlund-Nielsen, E.; Weirich, J.; Nørregaard, J.; Garnaes, J.; Mortensen, N. A.; Kris-tensen, A. Angle-independent structural colors of silicon. J. Nanophoton. , ,083988–083988.(18) Kumar, K.; Duan, H.; Hegde, R. S.; Koh, S. C.; Wei, J. N.; Yang, J. K. Printing colourat the optical diffraction limit. Nat. Nanotechnol. , , 557–561.(19) Roberts, A. S.; Pors, A.; Albrektsen, O.; Bozhevolnyi, S. I. Subwavelength plasmoniccolor printing protected for ambient use. Nano Lett. , , 783–787.(20) Ellenbogen, T.; Seo, K.; Crozier, K. B. Chromatic plasmonic polarizers for active visiblecolor filtering and polarimetry. Nano Lett. , , 1026–1031.(21) Diest, K.; Liberman, V.; Lennon, D. M.; Welander, P. B.; Rothschild, M. Aluminumplasmonics: optimization of plasmonic properties using liquid-prism-coupled ellipsom-etry. Opt. Express , , 28638–28650.(22) Li, Z.; Clark, A. W.; Cooper, J. M. Dual color plasmonic pixels create a polarizationcontrolled nano color palette. ACS Nano , , 492–498.1523) Duempelmann, L.; Casari, D.; Luu-Dinh, A.; Gallinet, B.; Novotny, L. Color renderingplasmonic aluminum substrates with angular symmetry breaking. ACS Nano , ,12383–12391.(24) Wang, L.; Ng, R. J. H.; Safari Dinachali, S.; Jalali, M.; Yu, Y.; Yang, J. K. Largearea plasmonic color palettes with expanded gamut using colloidal self-assembly. ACSPhotonics , , 627–633.(25) Richner, P.; Galliker, P.; Lendenmann, T.; Kress, S. J.; Kim, D. K.; Norris, D. J.;Poulikakos, D. Full-spectrum flexible color printing at the diffraction limit. ACS Pho-tonics , , 754–757.(26) Fan, J.; Wu, W.; Chen, Z.; Zhu, J.; Li, J. Three-dimensional cavity nanoantennaswith resonant-enhanced surface plasmons as dynamic color-tuning reflectors. Nanoscale ,(27) Ehrenreich, H.; Philipp, H. R. Optical properties of Ag and Cu.
Phys. Rev. , ,1622–1629.(28) West, P. R.; Ishii, S.; Naik, G. V.; Emani, N. K.; Shalaev, V. M.; Boltasseva, A.Searching for better plasmonic materials. Laser & Photonics Rev. , , 795–808.(29) Gérard, D.; Gray, S. K. Aluminium plasmonics. J. Phys. D: Appl. Phys. , ,184001.(30) Paniagua-Domínguez, R.; Yu, Y. F.; Miroschnichenko, A. E.; Krivitsky, L. A.; Fu, Y. H.;Valuckas, V.; Gonzaga, L.; Toh, Y. T.; Kay, A. Y. S.; Luk’yanchuk, B.; Kuznetsov, A. I.Generalized Brewster effect in dielectric metasurfaces. Nat. Commun. , , 10362.(31) Bonod, N. Silicon photonics: Large-scale dielectric metasurfaces. Nature Materials , , 664–665.(32) Jahani, S.; Jacob, Z. All-dielectric metamaterials. Nat. Nanotechnol. , , 23–36.1633) Decker, M.; Staude, I.; Falkner, M.; Dominguez, J.; Neshev, D. N.; Brener, I.;Pertsch, T.; Kivshar, Y. S. High-efficiency dielectric Huygens (cid:48) surfaces. Adv. Opt. Mat. , , 813–820.(34) Sautter, J.; Staude, I.; Decker, M.; Rusak, E.; Neshev, D. N.; Brener, I.; Kivshar, Y. S.Active tuning of all-dielectric metasurfaces. ACS Nano , , 4308–4315.(35) Li, J.; Verellen, N.; Vercruysse, D.; Bearda, T.; Lagae, L.; Van Dorpe, P. All-dielectricantenna wavelength router with bidirectional scattering of visible light. Nano Lett. , , 4396–4403.(36) Moitra, P.; Slovick, B. A.; Li, W.; Kravchencko, I. I.; Briggs, D. P.; Krishnamurthy, S.;Valentine, J. Large-scale all-dielectric metamaterial perfect reflectors. ACS Photonics , , 692–698.(37) Liu, S.; Sinclair, M. B.; Mahony, T. S.; Jun, Y. C.; Campione, S.; Ginn, J.; Ben-der, D. A.; Wendt, J. R.; Ihlefeld, J. F.; Clem, P. G.; Wright, J. B.; Brener, I. Opticalmagnetic mirrors without metals. Optica , , 250–256.(38) Hegde, R. S.; Panse, K. S. Design and optimization of ultrathin spectral filters basedon silicon nanocross antenna arrays. J. Nanophoto. , , 026030–026030.(39) Shalaev, M. I.; Sun, J.; Tsukernik, A.; Pandey, A.; Nikolskiy, K.; Litchinitser, N. M.High-efficiency all-dielectric metasurfaces for ultracompact beam manipulation in trans-mission mode. Nano Lett. , , 6261–6266.(40) Staude, I.; Miroshnichenko, A. E.; Decker, M.; Fofang, N. T.; Liu, S.; Gonzales, E.;Dominguez, J.; Luk, T. S.; Neshev, D. N.; Brener, I.; Kivshar, Y. Tailoring directionalscattering through magnetic and electric resonances in subwavelength silicon nanodisks. ACS Nano , , 7824–7832. 1741) Lin, D.; Fan, P.; Hasman, E.; Brongersma, M. L. Dielectric gradient metasurface opticalelements. Science , , 298–302.(42) Arbabi, A.; Horie, Y.; Bagheri, M.; Faraon, A. Dielectric metasurfaces for completecontrol of phase and polarization with subwavelength spatial resolution and high trans-mission. Nat. Nanotechnol. , 937–943.(43) Yue, W.; Gao, S.; Lee, S.-S.; Kim, E.-S.; Choi, D.-Y. Subtractive color filters based ona silicon-aluminum hybrid-nanodisk metasurface enabling enhanced color purity.
Sci.Rep. , , 29756.(44) Shrestha, V. R.; Lee, S.-S.; Kim, E.-S.; Choi, D.-Y. Polarization-tuned dynamic colorfilters incorporating a dielectric-loaded aluminum nanowire array. Sci. Rep. , ,12450.(45) Cao, L.; Fan, P.; Barnard, E. S.; Brown, A. M.; Brongersma, M. L. Tuning the color ofsilicon nanostructures. Nano Lett. , , 2649–2654.(46) Proust, J.; Bedu, F.; Gallas, B.; Ozerov, I.; Bonod, N. All-dielectric colored metasur-faces with silicon Mie resonators. ACS Nano , , 7761–7767.(47) Zhao, W.; Liu, B.; Jiang, H.; Song, J.; Pei, Y.; Jiang, Y. Full-color hologram usingspatial multiplexing of dielectric metasurface. Opt. Lett. , , 147–150.(48) Kuznetsov, A. I.; Miroshnichenko, A. E.; Fu, Y. H.; Zhang, J.; Luk’yanchuk, B. Mag-netic light. Sci. Rep. , , 00492.(49) .(50) http://colour.readthedocs.io/en/latest .(51) https://graphics.stanford.edu/courses/cs178/applets/custom-subtractive-too-wide.jpghttps://graphics.stanford.edu/courses/cs178/applets/custom-subtractive-too-wide.jpg