Nanowire lenses can perform two functions at once
Cameras can be found in an ever increasing number of consumer and biomedical applications. In fact, several cellular phones now on the market contain two cameras. As a consequence, there is a great demand to further shrink imaging systems without giving up the functionality found in much larger optical systems.
Frequently, much of an imaging system's functionality derives from features that are difficult to miniaturize, such as multiple lens elements and mechanical motion. Microscope objectives have multiple surfaces combined into a compound lens, while stereoscopic imaging systems require two lens apertures in order to collect a stereo pair. Both of these geometries add size and complexity to an imaging system. In addition, optical zoom requires the mechanical motion of one lens with respect to another, and consequently has been primarily neglected in miniaturized imaging systems.
In order to expand the functionality and maintain the simplicity of these systems, we have developed a lens technology that can capture two images of a scene without any mechanical motion and uses only a single lens.1 The lens can produce two different focusing functions depending on the illumination light's polarization. One image is produced for one polarization, and a completely independent image is produced for the perpendicular polarization. The lens characteristics can be switched either by controlling the illumination's polarization or by projecting both images at once using unpolarized light.
The polarization switching is based on the birefringent optical properties of elliptical cross-section silicon nanowires. The lens is composed of a large array of four different geometries of elliptical nanowires with different orientations and ellipticities: see Figure 1(a). By placing the nanowires across the array, we can encode two different lens functions into a single element, where each lens function is independently read out by a single polarization state. For each respective polarization, the lens acts as a two-level diffractive lens with a focal length and optical axis that can be designed into its function.
We implemented this lens technology into two optical systems. The first was a stereoscopic lens that can capture two object images from two different perspectives: see Figure 1(b). Typically, stereoscopic imaging systems use two different displaced lenses to generate the image pair. Here, it is derived from a single lens. In this case, the two different lens functions have the same focal length, but a shifted optical axis. The shift produces parallax, which can be used to reconstruct 3D information from the scene. Figure 1(c) and (d) shows a captured stereo pair of a target consisting of the letter ‘H,’ where the shift in the position of the ‘H’ is due to parallax, which depends on object depth.
The second imaging system we implemented was a lens that encoded two different focal lengths: see Figure 1(e). Tunable focal length lenses can be used either for optical zoom or to focus on an object at different depths. By precisely designing the two focal lengths, we implemented a lens capable of projecting two object images, each having a different magnification: see Figures 1(f) and (g).
As the integration of imaging systems grows increasingly more common and their dimensions continue to shrink, new lens technologies are required. Nanostructured materials allow light control based on color and polarization. Using these degrees of freedom, it will be possible to add more functionality to miniaturized systems. And using polarization, single lens elements can capture two independent images using the unpolarized nature of incoherent light. Recent advances in polarization-discriminating cameras will further advance these systems to provide new functionality.
Using color and polarization degrees of freedom, we hope to develop non-mechanical lens systems that can simultaneously collect more than two independent images from a scene at a single instance in time.
Ethan Schonbrun is a Junior Fellow at the Rowland Institute at Harvard. He is the principal investigator of the micro-optics and microfluidics group, which develops new methods to image cells in fluids. Previously, he received his PhD from the University of Colorado and completed a postdoctoral fellowship at Harvard University.
