Abstract
Planar structured interfaces, also known as metasurfaces, are continuously attracting interest owing to their ability to manipulate fundamental attributes of light, including angular momentum, phase, or polarization. However, chromatic aberration, limiting broadband operation, has remained a challenge for metasurfacesbased optical components and imagers. The limitation stems from the intrinsic dispersion of existing materials and design principles. Here we report and experimentally demonstrate polarizationindependent fishnetachromaticmetalenses with measured average efficiencies over 70% in the continuous band from the visible (640 nm) to the infrared (1200 nm). Results of the scalable platform are enabling for applications requiring broad bandwidth and high efficiency including energy harvesting, virtual reality and information processing devices, or medical imaging.
Introduction
In 1660, Newton published the discovery on the decomposition of white light by prisms and color theory. Since then, optical dispersion has continued to fascinate the scientific world. Dispersion is used in cosmology to measure the expansion of the universe via the propagation of gamma rays^{1}. It is also used in many optical applications including modelocked lasers and prism spectroscopy to cite a few, and, explains rainbows^{2}. However, dispersion is a major challenge for imaging systems^{3}. Chromatic aberration, originating from the variation of the refractive index of materials with frequency, limits the performance of broadband optical devices. The quintessential feature of chromatic aberration is the wavelengthdependent focal length leading to axial aberration that degrades the quality of images. To overcome this limitation, conventional opticalimaging systems often use an appropriate combination of lenses, such as the achromatic doublet or a combination of refractive and diffractive elements with opposite dispersion. Recently, interest has turned towards thinner, lighter devices known as metasurfaces^{4,5,6,7,8}. They are subwavelength nanostructured interfaces, capable of controlling optical waves. A large variety of components has been reported, including lenses^{9,10,11,12,13,14,15,16}, holograms, quarterwave plates, halfwave plates, vortex plates, carpet cloaks, concentrators, polarizers, thin absorbers, or sensors^{17,18,19,20,21,22,23,24,25,26,27,28,29,30,31}. Despite exciting findings, achieving simultaneously high efficiencies and large bandwidths has remained a challenge. Recent state of the art results in the visible reported a bandwidth from 470 to 670 nm with an efficiency of 20%^{14}, while in ref. ^{15} the efficiency of 40% was obtained for the bandwidth from 400 to 660 nm.
Here, we experimentally report polarizationindependent, fishnetachromaticmetalenses (FAM) with measured average efficiencies over 70% in the continuous band from the visible (640 nm) to the infrared (1200 nm). The design approach is based on the simultaneous control of the slope and the phaseshiftintercept, two parameters that need to be continuously optimized in the lens for achromatic operation. The lens is fabricated in TiO_{2} and experimental Strehl ratios larger than 80% are measured in the entire octave bandwidth demonstrating diffractionlimited operation.
Results
Conditions for broadband achromatic operation
In order to focus light to a point for a normal incident plane wave, a flat lens needs to deflect light by a position (r) dependent angle (θ) given by the relation^{5}:
where ϕ (r, f) is the phase profile required, f is the frequency, F is the focal length, r is the radial position, c is the speed of light, and g is a reference phase function independent of r. The reference phase can be an arbitrary function of frequency because only the spatial phase difference matters for the interference of waves at the same frequency after their interaction with the lens. We thus consider the phase shift, i.e., the phase difference between the local phase and the phase at the reference position taken at r = 0 (center of the lens). Hence, the phaseshift equation for a normal incident wave Δϕ(r, f) is
where m(r) is the frequency slope of the phaseshift. Equation (2) reveals the requirements of a broadband achromatic metalens. First, the phaseshift for all positions is linear with respect to frequency. This condition can be locally satisfied using waveguide modes. Second, the slope of the phaseshift (dispersion) varies with position following Eq. (2) and the phaseshift Δϕ(r, f) is proportional to frequency, i.e., the phaseshift intercept with respect to frequency is zero. The metasurface is thus a waveguide array with, ideally, a local and simultaneous control of the slope and the intercept of the phaseshift. To satisfy the requirements of achromatic broadband metalenses, we propose to use the crosscircle waveguide shown in Fig. 1a as building block. The building block has four geometrical parameters that are the radius (R), the width (W), the length (L), and the period (P). Constraints impose for example W ≤ 2R and L ≤ P. Using geometric parameters, the slope can be controlled with a quasicontrol of the phaseshift intercept consisting of minimizing it (ideally zero). Because each position has a unique (slope, phaseshift intercept) coordinate, dimensions can be chosen accordingly. One of the unique aspects of the device is that the design accounts for modified nearfield interactions that usually hinder the performance of metalenses as explained in ref. ^{23}. This is done via the isoslopes and isophaseshift intercepts used in the construction of our metasurfaces. It is important to note that the four geometric parameters are not independent, as a change in any of them can affect the effective index of the waveguide they form. This signifies that it is challenging to have perfect achromaticity and efficiency as phaseshift intercepts and slopes cannot be fully independently controlled in a planar design. The limitation confirms that this is intrinsically an optimization problem^{32,33,34,35}.
In metasurfaces, the spatial derivative of the slope controls the direction of incident rays to make them reach the focal point. It is thus important to have the correct slope to prevent chromatic effects and a decrease in efficiency. The intercept, however, controls the superposition of waves at the focal point, i.e., mostly affects the efficiency of the lens, not the position of the focal length. We can thus compromise on the intercept in the design of the lens. To quantify the impact of a nonzero phaseshift intercept on the efficiency of our metalens, Monte Carlo simulations are performed with 100 simulations for each element using a homemade finite difference time domain code. Each simulation was given a certain magnitude of the phaseshift intercept (error or deviation from the ideally zero phaseshift intercept) that was randomly distributed between unitcells. The focusing efficiency was then compared to the ideal metalens implementing not only the correct slope but also the correct phaseshift intercept. Results, presented in Supplementary information, indicate that an error on the phaseshift intercept smaller than 30° decreases the efficiency of the metalens by <10% and does not affect the position of the focal point.
Design of the FAMs
Figure 1a presents a sketch of the titanium dioxide (TiO_{2})based metasurface and the geometry of the unitcell with multiple degrees of freedom. It is a fishnetlike structure with a period P = 370 nm and a height H = 350 nm. The structure is fabricated by topdown nanomanufacturing methods and a scanning electron micrograph (SEM) of a fabricated metalens as shown in Fig. 1b. To design our metasurface, geometric parameters are controlled by pair, (W, R) in Fig. 2a, b and (L, R) in Fig. 2c, d. By considering fabrication limits, isoslopes and isophaseshift intercept plots of realistic geometries are computed using fullwave numerical simulations (CST Microwave Studio) and the local phase method^{23}, followed by leastsquare linear fitting. The phase shift of elements is calculated using a reference at the center of the lens with geometric parameters W = 270 nm, R = 135 nm, and L = P = 370 nm. For all other geometries, the parameters in Fig. 2 are calculated.
Figure 2a, c show that changes in R, W, and L enable slopes from zero to −0.35° THz^{−1} which in turn determines the maximum achievable size of the metalens for a given focal length. The figure also confirms that it is not possible to fully independently control the slope and the phase shiftintercept. However, accepting an error on the phase shift intercept enables designs sweeping all slope values. Figure 2a enables slopes from zero to −0.2° THz^{−1} while keeping a phaseshift intercept error below 30° (Fig. 2b). For the 20 μm × 20 μm metalens, we have chosen points indicated in blue (along the black arrow) to minimize discretization errors and points in the gray area are not geometrically allowed as W ≥ 2R. For absolute value of slopes larger than 0.2° THz^{−1}, we used parameters in Fig. 2c and the second trajectory (blue points along the black arrow) also keeps the phaseshift intercept error below 30° (Fig. 2d). The evolution of the geometry of the unitcell from the center of the lens to its edge is further discussed in supplementary information (Supplementary Fig. 1).
Fabrication of the FAMs
The designed structure is fabricated by topdown methods and the SEM of a metalens, presented in Fig. 1b, clearly shows the high quality of the implementation. The fabrication uses three major steps. The first step consists of patterning the polymethyl methacrylate (PMMA) resist using electron beam lithography (EBL) that is subsequently developed in solution to remove the exposed PMMA. The pattern is the inverse of our final metasurfaces. In the second step, the exposed sample is transferred to an atomic layer deposition (ALD). The ALD process deposits 350 nm of TiO_{2} so that all features are filled. The third step consists of removing the residual TiO_{2} film that coats the top surface of the resist using reactiveionetching. After removing PMMA, the TiO_{2} metasurfaces were obtained. It is worth noting that FAMs have mostly connected structures and are thus more stable mechanically than metasurfaces based on fully disconnected elements. Fabrication imperfections with a magnitude of ±5 nm decrease the efficiency by at most 8%, making the FAMs robust (Supplementary Fig. 10).
Characterization of the FAMs
The fabricated metalenses were optically characterized using a custom setup consisting of two main systems dedicated to illumination and imaging (Supplementary Fig. 5). The illumination system comprises a supercontinuum laser and an acoustooptic tunable filter to select the operating wavelength. For the imaging system, a ×50 extralong working distance microscope objective lens with a numerical aperture of 0.65 and a tube lens with a focal distance of 20 cm were used to image planes of interest on a camera. To image the focusing pattern, we moved the sample around the focal point using a translation stage.
Figure 3a presents the measured intensity profiles in the focal plane z = F (transverse x–y plane) of the metalenses at different wavelengths. The dots in Fig. 3b represent a normalized crosssection of the experimental measurements in Fig. 3b and the lines correspond to the theoretical Airy disk. Figure 3c shows the normalized intensity profiles in the plane y = 0 (axial x–z plane) around the focal point of the metalens at different wavelengths. Black circles represent the focal spots for different wavelengths. These results show nearly diffractionlimited focal spots with no obvious distortion. To further examine the performance of designed metalenses, we measured focal lengths, focusing efficiencies, and full widths at half maximum (FWHM) for different lens diameters as shown in Fig. 4. Figure 4a–c present SEM images of lenses of diameters 10, 15, and 20 μm. Figure 4d presents the focal length of the metalenses and shows that they are mostly unchanged when the wavelength varies from 640 to 1200 nm, demonstrating the successful realization of the broadband achromatic property. Figure 4e presents the focusing efficiency for metasurfaces of different diameters and focal lengths. To enable a quantitative comparison between our devices and previously reported metalenses, we defined the size of the focal spot as three times the FWHM in the measurements of the efficiencies^{9,16,36}. The measured efficiency of a metalens is defined as the focal spot power divided by the transmitted power through an aperture of the same diameter as the metalens. Efficiencies from 65% to 75% for the entire band are measured. In Fig. 4f (right axis), the experimental FWHM plots of the focal spots are presented. We also present in Fig. 4f (left axis) the Strehl ratio defined as the ratio of the peak focal spot irradiance of the manufactured FAMs to the focal spot irradiance of an aberrationfree lens. The calculation includes the total energy enclosed by the measured focal spot within a diameter up to the second dark ring of corresponding airy disk. These results show that FAMs successfully achieve a diffractionlimited focus. Similar results are obtained for the X and Y polarization confirming polarization independence (Supplementary Fig. 8). Beyond the extremities of the current bandwidth, the slopes are smaller than the target slopes at different positions owing to the dispersion of TiO_{2} and leading to an increased focal length. We have thus successfully implemented planar achromatic metalenses spanning the continuous wavelength range from 640 nm in the visible to 1200 nm in the infrared.
Discussion
To compare metasurfaces operating in various wavelengths range, a fair metric is the fractional bandwidth defined as the bandwidth divided by the central frequency, i.e., Δλ/λ_{center} = Δf/f_{center} with Δλ = λ_{max} − λ_{min} and λ_{center} = (λ_{max} + λ_{min})/2. Our FAMs have a fractional bandwidth of 61% with an efficiency of 70%. The fractional bandwidth in ref. ^{14} is 35% with an efficiency smaller than 20%. In ref. ^{15}, the fractional bandwidth is 41% with an efficiency smaller than 35%. To the best of our knowledge, the FAMs have higher efficiencies and larger fractional bandwidths than other experimentally reported metasurfaces. Moreover, compared to multilevel diffractive lenses presented in ref. ^{36}, the FAMs can be extended to anisotropic structures to enable functions not easily achieved with diffractive elements. Metalenses have the advantage to enable subwavelength unitcells which usually come at the price of the bandwidth and efficiency and this tradeoff is overcome in our design. Large scale metalenses are of technological importance. FAMs can be implemented at larger scale by increasing the maximum slope which will require higher aspect ratio. State of the art diffractive optics experiments have an efficiency of 35% and a fractional bandwidth of 62.7%^{37}. It is worth noting that metalenses have larger angular transmission compared to Fresnel lenses, which suffer from the shadowing effect due to their sawtooth surface profile^{38}. Our work thus brings metasurfaces to a performance level not previously reached.
We proposed and experimentally demonstrated metalenses combining high efficiency, polarization independence, and achromaticity in the continuous wavelength range from 640 nm in the visible to 1200 nm in the infrared. The broadband operation is achieved by enforcing the slopes of the phaseshift that vary continuously from the center of the lens to its edge, and, by minimizing the phaseshift intercepts that are ideally zero for achromatic operation. To the best of our knowledge, this is the broadest band achromatic metalens reported to date. The proposed approach significantly extends the current state of the art of metalenses both in terms of bandwidth and efficiency and opens the door to many applications.
Methods
Numerical simulations
Numerical simulations are performed using finitelement method (CST Microwave Studio) and the local phase method, followed by leastsquare linear fitting. The phase shift of elements is calculated using a reference at the center of the lens with geometric parameters W = 270 nm, R = 135 nm, and L = P = 370 nm.
Sample fabrication
The fabrication flow chart is shown in Supplementary Fig. 3. An electron beam resist (EBR) [370 nm PMMA A4] is coated on the sample (1500 rpm) followed by 1 min bake at 180 °C (hot plate). The metasurface pattern is written in the resist using EBL. We patterned the resist using EBL followed by a development process to remove the exposed EBR. This resist pattern is the inverse of our final metasurface. We transferred the patterned sample to an ALD chamber at low temperature. The lowtemperature deposition is important to obtain the desired amorphous material and to avoid the contamination of the ALD chamber. Using planarization, we removed the residual TiO_{2} film on top of the surface of the resist by reactiveionetching using a mixture of BCl_{3} and Cl_{2} gases. The etch depth was equal to t_{film} so that the etching process exposes the underlying resist and the top of the nanostructures. We finally removed the remaining EBR and left only the TiO_{2} metasurfaces.
Optical measurements
To optically characterize the fabricated lenses, a custom setup was used. The experimental setup (Supplementary Fig. 5) is composed of two main systems dedicated to illumination and imaging. The illumination system comprises a supercontinuum laser (NKT photonics) and an acoustooptic tunable filter (Super K) to select the operating wavelength from 640 to 1200 nm illuminating 100% of the lens surface. For the imaging system, a ×50 extralong working distance microscope objective lens with a NA = 0.65 and a tube lens with a focal distance of 20 cm were used to image intensity at planes of interest with a camera. To image different planes of the sample, we moved the sample using a translation stage around the focal point.
Data availability
The data that support the plots within this paper and other findings of this study are available from the corresponding author upon reasonable request.
Code availability
The computer codes that support the plots within this paper and other findings of this study are available from the corresponding author upon reasonable request.
References
 1.
Hubble, E. A relation between distance and radial velocity among extragalactic nebulae. Proc. Natl Acad. Sci. USA 15, 168–173 (1929).
 2.
Ball, D. W. The Basics of Spectroscopy. (SPIE Press, Bellingham, Washington, 2001).
 3.
Herzberger, M. & McClure, N. R. The design of superachromatic lenses. Appl. Opt. 2, 553–560 (1963).
 4.
Lalanne, P., Astilean, S., Chavel, P., Cambril, E. & Launois, H. Blazed binary subwavelength gratings with efficiencies larger than those of conventional échelette gratings. Opt. Lett. 23, 1081–1083 (1998).
 5.
Bomzon, Z., Kleiner, V. & Hasman, E. PancharatnamBerry phase in space variant polarizationstate manipulations with subwavelength gratings. Opt. Lett. 26, 1424–1426 (2001).
 6.
Lu, F., Sedgwick, F. G., Karagodsky, V., Chase, C. & ChangHasnain, C. J. Planar highnumericalaperture lowloss focusing reflectors and lenses using subwavelength high contrast gratings. Opt. Express 18, 12606–12614 (2010).
 7.
Fong, B. H., Colburn, J. S., Ottusch, J. J., Visher, J. L. & Sievenpiper, D. F. Scalar and tensor holographic artificial impedance surfaces. IEEE Trans. Antennas Propag. 58, 3212–3221 (2010).
 8.
Yu, N. F. et al. Light propagation with phase discontinuities: generalized laws of reflection and refraction. Science 334, 6054 (2011).
 9.
Arbabi, A., Horie, Y., Ball, A. J., Bagheri, M. & Faraon, A. Subwavelengththick lenses with high numerical apertures and large efficiency based on high contrast transmit arrays. Nat. Commun. 6, 7069 (2015).
 10.
Khorasaninejad, M. et al. Metalenses at visible wavelengths: diffractionlimited focusing and subwavelength resolution imaging. Science 352, 6290 (2016).
 11.
Pors, A., Nielsen, M. G., Eriksen, R. L. & Bozhevolnyi, S. I. Broadband focusing flat mirrors based on plasmonic gradient metasurfaces. Nano Lett. 13, 829–834 (2013).
 12.
Khorasaninejad, M. et al. Achromatic metalens over 60 nm bandwidth in the visible and metalens with reverse chromatic dispersion. Nano Lett. 17, 1819–1824 (2017).
 13.
Wang, S. M. et al. Broadband achromatic optical metasurface devices. Nat. Commun. 8, 187 (2017).
 14.
Chen, W. T. et al. A broadband achromatic metalens for focusing and imaging in the visible. Nat. Nanotechnol. 13, 220–226 (2018).
 15.
Wang, S. M. et al. A broadband achromatic metalens in the visible. Nat. Nanotechnol. 13, 227–232 (2018).
 16.
Shrestha, S., Overvig, A. C., Lu, M., Stein, A. & Yu, N. Broadband achromatic dielectric metalenses. Light: Sci. Appl. 7, 85 (2018).
 17.
Chen, W. T., Zhu, A. Y., Sisler, J., Bharwani, Z. & Capasso, F. A broadband achromatic polarizationinsensitive metalens consisting of anisotropic nanostructures. Nat. Commun. 10, 355 (2019).
 18.
Zheng, G. X. et al. Metasurface holograms reaching 80% efficiency. Nat. Nanotechnol. 10, 308–312 (2015).
 19.
Yu, N. F. et al. A broadband, backgroundfree quarterwave plate based on plasmonic metasurfaces. Nano Lett. 12, 6328–6333 (2012).
 20.
Kats, M. A., Blanchard, R., Genevet, P. & Capasso, F. Nanometre optical coatings based on strong interference effects in highly absorbing media. Nat. Mater. 12, 20 (2013).
 21.
Hsu, L. Y., Lepetit, T. & Kanté, B. Extremely thin dielectric metasurface for carpet cloaking. Prog. Electromagn. Res. 152, 33–40 (2015).
 22.
Ni, X., Wong, Z. J., Mrejen, M., Wang, Y. & Zhang, X. An ultrathin invisibility skin cloak for visible light. Science 349, 6254 (2015).
 23.
Hsu, L., Dupré, M., Ndao, A., Yellowhair, J. & Kanté, B. Local phase method for designing and optimizing metasurface devices. Opt. Express 25, 24974–24982 (2017).
 24.
Devlin, R. C., Ambrosio, A., Rubin, N. A., Mueller, J. P. B. & Capasso, F. Arbitrary spintoorbital angular momentum conversion of light. Science 358, 6365 (2017).
 25.
Ding, Q. et al. Feasibility analysis of nanostructured planar focusing collectors for concentrating solar power applications. ACS Appl. Energy Mater. 12, 6927–6935 (2018).
 26.
Shaltout, A. M., Shalaev, V. M. & Brongersma, M. L. Spatiotemporal light control with active metasurfaces. Science 364, 6441 (2019).
 27.
Divitt et al. Ultrafast optical pulse shaping using dielectric metasurfaces. Science 364, 6443 (2019).
 28.
Ren, H. et al. Metasurface orbital angular momentum holography. Nat. Commun. 10, 2986 (2019).
 29.
Shaltout, A. M. et al. Spatiotemporal light control with frequencygradient metasurfaces. Science 365, 6451 (2019).
 30.
Ding, Q. et al. Feasibility analysis of nanostructured planar focusing collectors for concentrating solar power applications. ACS Appl. Energy Mater. 1, 6927–6935 (2018).
 31.
Zhou, Y. et al. Multifunctional metaoptics based on bilayer metasurfaces. Light: Sci. Appl. 8, 80 (2019).
 32.
Phan, T. et al. Highefficiency, largearea, topology optimized metasurfaces. Light: Sci. Appl. 8, 48 (2019).
 33.
Karagodsky, V., Sedgwick, F. G. & ChangHasnain, C. J. Theoretical analysis of subwavelength high contrast grating reflectors. Opt. Express 18, 16973–16988 (2010).
 34.
Xiao, T. P. et al. Diffractive spectralsplitting optical element designed by adjointbased electromagnetic optimization and fabricated by femtosecond 3D direct laser writing. ACS Photonics 3, 886–894 (2016).
 35.
Zhang, H., Hsu, C. W. & Miller, O. D. Scattering concentration bounds: brightness theorems for waves. Optica 10, 1321–1327 (2019).
 36.
Banerji, S. et al. Imaging with flat optics: metalenses or diffractive lenses? Optica 6, 805–810 (2019).
 37.
Monjurul, M., Majumder, A. & Menon, R. Multiplane, multiband image projection via broadband diffractive optics. Appl. Opt. 1, 38–44 (2020).
 38.
Lalanne, P. & Chavel, P. Metalenses at visible wavelengths: past, present, perspectives. Laser Photon. Rev. 11, 1600295 (2017).
Acknowledgements
This research was mostly supported by the US Department of Energy (DOE) (grant no. DEEE0007341). It was partially supported by the National Science Foundation Career Award (ECCS1554021), the Office of Naval Research Young Investigator Award (N000141712671), the ONR JTO MRI Award (N000141712442), and the DARPA DSONLM Program no. HR00111820038. The work was performed in part at the San Diego Nanotechnology Infrastructure, a member of the National Nanotechnology Coordinated Infrastructure, which is supported by the National Science Foundation (grant no. ECCS1542148). We thank M. Montero for technical assistance regarding the fabrication.
Author information
Affiliations
Contributions
B.K. conceived the project. A.N. and L.H. designed the structures. L.H. performed the numerical calculations. J.H. performed the fabrication with J.H.P. A.N. performed the measurements. A.N., C.C.H., and B.K. supervised the project. All authors contributed to data analysis, discussions, and manuscript writing.
Corresponding author
Ethics declarations
Competing interests
All authors declare competing interests. The Regents of the University of California have filed a patent application (US Prov. App. No. 62/941,077) on technology related to the processes described in this article.
Additional information
Peer review information Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Ndao, A., Hsu, L., Ha, J. et al. Octave bandwidth photonic fishnetachromaticmetalens. Nat Commun 11, 3205 (2020). https://doi.org/10.1038/s41467020170159
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467020170159
Further reading

Highperformance gallium nitride dielectric metalenses for imaging in the visible
Scientific Reports (2021)

Highefficiency broadband achromatic metalens for nearIR biological imaging window
Nature Communications (2021)

Neural nanooptics for highquality thin lens imaging
Nature Communications (2021)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.