Creating arbitrary illusions by planar ultrathin metasurfaces

Abstract

Many interesting wave manipulation devices, including illusion devices, have been proposed and realized by transformation optics. Using illusion devices, an object of arbitrary shape can be made to appear like another object. However, there are substantial limitations in current illusion devices based on transformation optics. Firstly, refractive index must be modulated over a large volume, thus, the dimension of current illusion devices cannot be small. Moreover, sophisticated three-dimensional structures are required and difficult to be fabricated. Here, a method to create arbitrary illusions using planar ultrathin metasurfaces is proposed, as shown in Figure 1. Figure 1(a) shows a surface with arbitrary shape, generated by the Peaks function in Matlab. The length and width of the Peaks surface are both 12λ and the height is between −0.5λ and 0.5λ, where λ is the free-space wavelength of our referenced frequency, 5.8 GHz. The Peaks surface in Figure 1(a) is pretended with a planar metasurface, as shown in Figure 1(b), by creating the similar field distribution. At the point on the planar metasurface corresponding to an arbitrary point on the Peaks surface with height h, the metasurface cell should recover the phase of the scattered wave. The incident wave at an oblique angle (θ, φ) at that point should scatter as if it were reflected by the Peaks surface The metasurface cell at that point should provide a phase shift φ = π+2k 0 h cos θ to compensate the phase difference between the Peaks surface and planar illusive metasurface. When the Peaks surface and the planar illusive metasurface are illuminated from z direction (θ = 0°) by a planar wave, the interferential electric fields of incident wave and reflective wave in z = λ plane are respectively shown in Figures 1(c) and (d), two similar distributions. Therefore, the Peaks surface in Figure 1(a) is successfully pretended by the planar illusive metasurface in Figure 1(b), which demonstrates that our method is feasible.