Light Structuring & Manipulation:
Metasurfaces, two dimensional (2D) metamaterials comprised of subwavelength features, can be used to tailor the amplitude, phase, and polarisation of an incident electromagnetic wave propagating at an interface. Though many novel metasurfaces have been explored, the hunt for cost-effective, highly efficient, low-loss and polarisation insensitive applications is ongoing. In this work, we utilize an efficient and cost-effective dielectric material, hydrogenated amorphous silicon (a–Si:H), to create an ultra-thin transmissive surface that simultaneously controls phase. This material exhibits significantly lower absorption in the visible regime compared to standard amorphous silicon, making it an ideal candidate for various on-chip applications. Our proposed design, which works on the principle of index waveguiding, integrates two distinct phase profiles, that of a lens and of a helical beam, and is versatile due to its polarisation-insensitivity. We show how this metasurface can lead to highly concentrated optical vortices in the visible domain, whose focused ring-shaped profiles carry orbital angular momentum at the miniaturized scale.
Optical Holography & Color Filtering:
Dielectric materials that are low-loss in the visible spectrum provide a promising platform to realize the pragmatic features of metasurfaces. Here, all-dielectric, highly efficient, spin-encoded transmission-type metaholograms (in the visible domain) are demonstrated by utilizing hydrogenated amorphous silicon (a-Si:H). In comparison to previously reported visible metaholograms based on TiO2 and other dielectric materials, all-dielectric metasurfaces provide a cost-effective more straightforwardly fabricated (aspect ratio 4.7), CMOS compatible, and comparably efficient solution in the visible domain. A unique way of utilizing polarization as an extra degree of freedom in the design to implement transmission-type helicity-encoded metaholograms is also proposed. The produced images exhibit high fidelity under both right and left circularly polarized illuminations. The proposed cost-effective and CMOS-compatible material and methods open up an avenue for on-chip development of numerous new phenomena with high efficiency in the visible domain.
Meta-lenses for Compact Imaging:
Plasmonic & Dielectric Meta-sensing:
Food contamination (particularly in daily edible items like milk, oil, yogurt, etc.) is a common practice in developing countries to make more money and to meet the excessive demand (with low-cost). Since most of such adulterations are made by illiterate people who are unaware of hazardous health effects associated with such contaminations, consuming such impure food can result in various dangerous health effects. Though government regulatory authorities have detection mechanisms, they are not only expensive (with general spectroscopic analysis) but also inaccessible for domestic users for daily usage. Owing to such a situation, there is a definite need for a portable and straightforward detection mechanism for domestic users to avoid serious health issues. In this paper, we present all-dielectric meta-biosensors to identify (both qualitatively and quantitatively) the hazardous adulterants in water and milk via refractive index sensing. Our demonstrated platform is a highly efficient and unique way of testing water and milk with numerically measured sensitivity of ~ 500nm/RIU, a figure of merit (FoM) of ~2.5 and response of ~8 (800 %) change in transmittance per refractive index unit. The proposed transmittance based single wavelength amplitude measurement could significantly reduce the cost and the size of sensing devices without compromising sensitivity. Our demonstrated silicon-based design provide a route to realizing various highly efficient and affordable sensors for qualitative and quantitative analysis of different food contaminants.
Deep-learning & Quantum Enabled Meta-devices
Metasurfaces are 2D subwavelength devices which are used in various applications due to their unique ability to manipulate electromagnetic waves in the microwave and optical frequencies. Regardless of the rapid development of metasurfaces in recent years, it suffers from a major issue that is it’s extremely time-consuming and computational resource consuming design procedure. A lot of manual work, numerical solutions and multiple simulations to achieve metasurfaces having desired responses. In the more recent past, a novel concept of digital coding metamaterials was introduced which results in programmable metasurfaces. Meta-atoms of such metasurfaces are represented by binary codes which makes it possible to design them intelligently via deep learning networks. The use of deep learning neural networks have come out as an effective solution to predict the best possible designs and geometrical parameters of the nanostructure or the unit cells. This project aims to provide a deep learning neural network to design metasurfaces. This model will help to discover the non-predictable relationship between a metamaterial structure and its optical responses from a number of training examples, which circumvents the time-consuming, case-by-case numerical simulations in conventional metamaterial designs.