High-throughput super-resolution hyperlens imaging
Super-resolution imaging is of prime importance for nanoscience. However, the maximum resolution of a conventional optical system is determined by Abbe's diffraction limit,1 i.e., at half the wavelength of the light employed. To overcome this limitation, it is necessary to transfer both evanescent waves—which decay exponentially in naturally occurring materials—and propagating waves. Various super-resolution imaging methods take this approach, including near-field scanning optical microscopy,2,3 stochastic optical reconstruction microscopy,4 and stimulated emission depletion microscopy.5 However, these methods can be difficult to control and have serious noise problems because they obtain evanescent waves directly from the near field using a detector. More recently, metamaterials—artificially structured materials whose optical properties are determined by the design of subwavelength building blocks—have also been suggested. Whereas conventional materials have properties originating from a particular combination of elements and their arrangement, metamaterials have desired optical properties to support evanescent waves as a result of the specific design of their unit cells.
One interesting branch of this technology is hyperbolic metamaterials, which behave like both metal and insulator simultaneously. Hyperbolic metamaterials offer new possibilities for far-field super-resolution imaging, while eliminating the need for complex optical systems or algorithms. Hyperbolic metamaterials used for super-resolution have a multilayer structure, in which metals and dielectrics with subwavelength thickness are stacked alternately in a curved geometry. Incident electromagnetic waves cause the electron density to oscillate collectively along the metal-dielectric interfaces as a result of the subwavelength geometry. Therefore, the evanescent waves are supported because oscillation only occurs on the metal surface. Imaging using this approach has been demonstrated both theoretically6, 7 and experimentally.8
A hyperlens—a lens made of hyperbolic metamaterials—has cylindrical geometry, and therefore magnifies an image in one dimension. We showed, through both simulation and experiment, that a spherical hyperlens can provide super-resolution imaging in two lateral directions in the visible frequency range. Our hyperlens—shown in Figure 1(a)—comprises silver and titanium oxide, and achieved resolution of 160nm with a magnification ratio of 2.08: see Figure 1(b).9 Although the hyperlens is an attractive platform for far-field super-resolution imaging of living specimens in real time, its curved geometry, inefficient fabrication processes, and difficulties in aligning the object on the hyperlens make its application challenging. To overcome these issues, we introduced a quick and simple large-area fabrication process by adopting a hexagonal hyperlens array.10
Figure 2(a) shows a schematic of the setup for the hyperlens imaging system. The incident light is collimated with a wavelength of around 410nm. To avoid the difficulties in positioning the sample on the hyperlens, we fabricated a flat plate with a hexagonal hyperlens array: see Figure 2(b). Furthermore, we reduced the time and cost required to fabricate the chrome mask pattern by using a nanoimprinting method with chrome etching and a lift-off technique, instead of the conventional electron beam lithography. We deposited alternate layers of metal and dielectric material on every hemisphere using an electron beam evaporation process. Figure 2(c) shows a scanning electron microscopy image of the fabricated hyperlens array.
To make the chrome mask pattern, we deposited chrome, a sacrificial layer, and hydrogen silsesquioxane on a refined quartz wafer, and—as a master stamp—we nanoimprinted dense hemispherical patterns on top of the layered material using a patterned sapphire substrate with a diameter larger than 1.5μm and pitch greater than 1.5μm. We then etched the lower layers using reactive-ion etching and inductively coupled plasma etching. We followed this with a lift-off technique so that holes formed in a hexagonal pattern in the chrome layer. The chrome mask pattern was then formed on the quartz.
We defined hemispherical geometry on the quartz by isotropic wet etching with 10:1 buffered oxide etchant, and removed the chrome mask layer by wet etching using CR-7 etchant. Finally, we deposited alternately five pairs of silver and titanium oxide lenses, which had thicknesses of 15nm, using electron beam evaporation. To minimize the surface roughness, we maintained the vacuum conditions at 10−7 torr and the film growth rate at 0.1nm/s. Resolution depends on the thickness of each layer. Therefore, to improve performance, we examined the lens diameter and layer number using simulations. We improved the resolution and magnification ratio by increasing the ratio between the outer and inner radii of the hyperlens (although, as a result, the transmission—or intensity—of the image was reduced). Using our hexagonal hyperlens array, we were able to obtain images of living cells, such as neurons.
In contrast with other super-resolution imaging methods, our flat hexagonal hyperlens array can be produced quickly using large-area fabrication. The technique transfers evanescent waves to the far field without involving difficult fabrication or sample alignment processes. Therefore, hexagonal hyperlens arrays will be a good solution to super-resolution imaging of living samples, such as single molecules, neurons, and DNA. Our approach will enable the observation of cell processes, such as growth, disease, and information transfer, and will have potential applications in biotechnology, medical science, and industry. Our next steps will be to develop a reliable packaging for the hyperlens component, enabling a platform for super-resolution imaging that can be used like a filter in an optical microscope.
Minkyung Kim and Dasol Lee are graduate students in the Department of Mechanical Engineering.
Junsuk Rho is an assistant professor in the Department of Mechanical Engineering and the Department of Chemical Engineering. He received his BS, MS, and PhD in mechanical engineering at Seoul National University, the University of Illinois at Urbana-Champaign, and the University of California at Berkeley, respectively.