We report here the designing of optically-nonactive metamaterial by assembling metallic helices with different chirality. With linearly polarized incident light, pure electric or magnetic resonance can be selectively realized, which leads to negative permittivity or negative permeability accordingly. Further, we show that pure electric or magnetic resonance can be interchanged at the same frequency band by merely changing the polarization of incident light for 90 degrees. This design demonstrates a unique approach to construct metamaterial.

A model is developed to deal with lateral growth of a crystalline layer on a foreign substrate, which is mediated by successive nucleation at the concave corner defined by the meeting of a crystal facet and the substrate. It is demonstrated that due to an imbalance of surface/interface tensions at the concave corner, once the embryo of a nucleus is formed, the crystallographic orientation of the nucleus is spontaneously twisted. By successive nucleation at the concave corner, the crystalline layer develops laterally on the substrate, with its crystallographic orientation continuously rotated. In this way, a regular spatial pattern with well-defined long-range order is eventually achieved. Our model provides a criterion to predict when such an effect becomes observable in the nucleation-mediated lateral growth. The theoretical expectations are consistent with the experimental observations.Thin film growth has been extensively investigated in recent decades. Previous studies concentrated mostly on vertical growth (i.e., on the increase of film thickness above a substrate). 1,2 To initiate a thin film growth, nevertheless, horizontal extension of individual crystalline islands on the substrate is an important step. It is known that the crystallographic orientation of an epitaxial film is well aligned with that of the substrate. Yet in some cases the crystallographic orientation of the film may become complicated. [3][4][5] One interesting scenario is that nucleation selectively occurs at the concave corner of an expanding crystalline island and the substrate, and the island expansion (hence the thin film growth) is a repeated nucleation process at the concave corner with the orientation of each new nucleus heavily influenced by the surface tensions at the corner site. Once an embryo of nucleus (nascent nucleus) forms at the concave corner, the asymmetric local surface tensions will apply a torque to the embryo. 6 Hence, the crystallographic orientation is rotated with respect to the previous nucleus. Indeed, it has been observed that, during lateral growth of NH 4 Cl crystallite on a glass plate, the crystallographic orientation is continuously rotated, leading to either a periodic distribution of faceted regions and roughened regions 7 or a regular zigzag branches. 8,9 The appearance of the specific morphology depends on the index of the initial crystalline facet contacting the substrate and the axis of rotation. 9 Another example is the crystallographic wing tilt commonly observed in lateral overgrowth, 10,11 which is generally ascribed to the substrate effect. The underlying mechanisms of these lateral-growth-associated phenomena, however, are still not very clear, and the relationship between the rotation of the crystallographic orientation and the longrange order in aggregating crystallites has yet to be understood. In this paper, we try to establish fundamental relationships between nucleation and the rotation of crystallographic orientation. Based on previous experimental observations, we propose a model to...

We study the electro-optic (E-O) properties of a BaTiO 3 thin layer placed in a stack of dielectric layers, including a subwavelength diffraction grating with a two-dimensional periodicity, aiming to tune spectrally the position of the resonant reflection peak that is used for narrowband optical filtering. BaTiO 3 is chosen due to its strong E-O properties. When an external electric field is applied to the E-O layer, it leads to a spectral shift of the resonant peak. We study numerically different configurations with either weak or strong spectral tunability, presenting some arguments to explain these different behaviors. Taking into account only the linear part of the E-O effect (Pockels effect), the tuning of the peak that has 0.1 nm spectral width is approximately 33 nm for a 1.5 × 10 7 V∕m applied field. The shift is multiplied by three (97 nm) when also taking into account the quadratic E-O effect.