Optical metamaterials are usually based on planarized, complex-shaped, resonant nano-inclusions. Three-dimensional geometries may provide a wider set of functionalities, including broadband chirality to manipulate circular polarization at the nanoscale, but their fabrication becomes challenging as their dimensions get smaller. Here we introduce a new paradigm for the realization of optical metamaterials, showing that three-dimensional effects may be obtained without complicated inclusions, but instead by tailoring the relative orientation within the lattice. We apply this concept to realize planarized, broadband bianisotropic metamaterials as stacked nanorod arrays with a tailored rotational twist. Because of the coupling among closely spaced twisted plasmonic metasurfaces, metamaterials realized with conventional lithography may effectively operate as three-dimensional helical structures with broadband bianisotropic optical response. The proposed concept is also shown to relax alignment requirements common in three-dimensional metamaterial designs. The realized sample constitutes an ultrathin, broadband circular polarizer that may be directly integrated within nanophotonic systems.
We apply the Pancharatnam-Berry phase approach to plasmonic metasurfaces loaded by highly nonlinear multiquantum-well substrates, establishing a platform to control the nonlinear wave front at will based on giant localized nonlinear effects. We apply this approach to design flat nonlinear metasurfaces for efficient second-harmonic radiation, including beam steering, focusing, and polarization manipulation. Our findings open a new direction for nonlinear optics, in which phase matching issues are relaxed, and an unprecedented level of local wave front control is achieved over thin devices with giant nonlinear responses.
Optical activity in sum-frequency vibrational spectra has been observed for the first time in chiral liquids. The electric-dipole allowed chiral element of the nonlinear susceptibility appears to be 3 orders of magnitude smaller than typical allowed achiral elements. This is partly because the observed chirality requires a breakdown of the Born-Oppenheimer approximation.