Metasurfaces are optical elements with nanoscale dimensions and planar profiles that expand the functionality and usefulness of traditional refractive optics. This has lead to considerable interest in metasurfaces for a variety of applications, including consumer electronics, automotive, and medical devices. Metasurface elements can be manufactured using standard microelectronics process techniques enabling mass production of optical elements in semiconductor foundries. Additionally, being produced at the wafer-scale, metasurfaces enable wafer-scale integration of optical systems. However, most metasurfaces to date have been produced in small-scale research settings using serial patterning techniques such as electron beam lithography. In order for metasurfaces to penetrate high volume applications, scalable production processes must be developed, including lithographic reproduction of metasurface design geometries. In particular, the constituent elements of metasurfaces, nanopillars, have shapes, aspect ratios, and local density variations which diverge from typical microelectronics design rules.
After a brief overview of metasurface fundamentals and their applications we will discuss traditional fabrication techniques, device requirements, and the challenges that arise when scaling to manufacturing. We will also discuss future generations of metasurfaces and further challenges in terms of geometry, critical dimensions, and materials compatibility.
In polarimetry, that is, a measurement of the four-component polarization Stokes vector, a measurement must either consist of four (or more) sequential intensity measurements, sacrificing time resolution, or contain four separate light paths each with separate polarization optics, increasing bulk, cost, and system complexity. Similar issues present difficulty across polarization optics technology.
Metasurfaces, nanophotonic arrays of phase shifting elements, have emerged as a novel platform for polarization optics. These individual phase shifters can be designed with a characteristic anisotropy, and are thus imbued with tunable shape birefringence. A metasurface, then, can function as a subwavelength spaced array of nanoscale waveplates.
I will describe how, through relatively simple optimization methods, a metasurface producing arbitrarily specified polarization states (when illuminated with light of a known polarization) can be designed. This functionality is equivalent to a traditional diffraction grating with individual waveplate optics on each order; here, all the necessary polarization optics can be integrated into a flat, ultrathin optical element. Moreover, such a metasurface can be used in a reverse configuration as a parallel snapshot polarimeter with no need for additional polarization optics (save for a single polarizer). I present a detailed experimental characterization of both concepts in the visible spectral region and a comparison of the performance of the metasurface to a commercially available rotating waveplate polarimeter. With no bulk birefringent crystal optics, a parallel, full-polarization state measurement can be made with an integrated, scalable, and inexpensive device. Given its diffractive nature, the design naturally extends to spectropolarimetry and polarization imaging.
Optical elements that couple the spin/orbital angular momentum (SAM/OAM) of light have found a range of applications in classical and quantum optics. The J-plate, which refers to the variable denoting the photon’s total angular momentum (TAM), is a metasurface device that allows converting arbitrary, orthogonal input SAM states into two unique OAM states. Using independent phase control of any orthogonal basis of polarization states, the J-plate permits the conversion of arbitrary polarizations into states with arbitrary OAM. Here, we present a further development: Cascaded J-plates provide for versatile combinations of OAM states on any orthogonal basis of spin states. J-plates operating on different polarization bases and imparting independent values of OAM are designed and experimentally demonstrated to generate multiple OAM channels with different polarization states. The generated OAM states are determined by the superposition of the OAM states of the individual J-plates while the generated SAM states are determined by the polarization basis of the last J-plate. Theoretically, there are maximum of 2^n channels of OAM and n×2^n channels of TAM that can be generated by n such cascaded J-plates. It is also demonstrated that cascaded J-plates may produce complex structured light. Cascading J-plates provides a new way to control the TAM of a laser beam. These results may find application in quantum and classical communication.
We present a new platform that realizes high performance metasurfaces in the visible spectrum. This platform is based on atomic layer deposition of titanium dioxide and allows molding incident light wavefront to desired shapes including holographic images, optical vortices, and Bessel beams. The focus of this work will be on the design and demonstration of planar metalenses. We report on our recent experimental realization of high numerical aperture metalenses with efficiency as high as 86%. These metalenses can focus light into a diffraction-limited spot and can be employed for imaging purposes to provide sub-wavelength imaging resolution. In addition, by the judicious design of metalens building blocks, one can achieve a multispectral chiral metalens (MCML) within a single metasurface layer. The MCML can simultaneously resolve chiral and spectral information of an object without the requirement of additional optical components such as polarizers, wave-plates, or even gratings. Using this MCML, we map the chiroptical properties of a macroscopic chiral biological specimen across the visible range. Finally, since many applications require polarization insensitive planar lenses, we discuss the experimental realization of such metalenses with numerical apertures as high as NA=0.85. These metalenses can focus incident light to a spot as small as ~0.6lambda with efficiencies up to 70%. The straightforward and CMOS-compatible fabrication process of this platform is promising for a wide range of optics-based applications in multidisciplinary science and technology.
Using immersion lenses is a common approach to enhance the resolving power in various fields of optics such as microscopy and lithography. However, conventional immersion lenses are bulky, high-cost and are typically designed for only a few specific immersion liquids. The development of meta-surfaces provides a promising approach to manipulate light in a compact configuration, enabling many optical devices such as polarizers, waveplates and lenses. These are mainly focused in the near-infrared or the long-wavelength region of the visible spectrum due to fabrication challenges and intrinsic losses of materials used. Here, we demonstrate oil immersion planar lenses with a numerical aperture of 1.1 at visible wavelengths. The lenses provide diffraction-limited focal spots with Strehl ratios higher than 0.9 and 0.8 at their design wavelengths of 532 nm and 405 nm, respectively. Fabrication is based on an atomic-layer deposition (ALD) of TiO2. The loss of TiO2 in the visible is negligible and the surface roughness is well-controlled due to the precise monolayer growth of the TiO2 film. By applying the lens (designed at 532 nm) in a confocal scanning microscopy setup, we are able to achieve high-quality images with sub-wavelength resolution. It should be noted that this lens can be efficiently tailored for any liquid. We demonstrate another design for water-immersion lenses, which are highly applicable to super-resolution bio-imaging applications. The compactness and design flexibility of this platform is highly promising for widespread applications in imaging and spectroscopy.