The fovea centralis is a closely-packed vertical array of inverted-cone photoreceptor cells located in the retina that is responsible for high acuity binocular vision. The cones are operational in well-lit environments and are responsible for trapping the impinging illumination. We present the vertical light-funnel silicon array as a light-trapping technique for photovoltaic applications that is bio-inspired by the properties of the fovea centralis. We use opto-electronic simulations to evaluate the performance of light-funnel solar cell arrays. Light-funnel arrays present ~65% absorption enhancement compared to a silicon film of identical thickness and exhibit power conversion efficiencies that are 60% higher than those of optimized nanowire arrays of the same thickness although nanowire arrays consist of more than 2.3 times the amount of silicon. We demonstrate the superior absorption of the light-funnel arrays as compared with recent advancements in the field. Fabrication of silicon light-funnel arrays using low-cost processing techniques is demonstrated.

We report a specific, label-free and real-time detection of femtomolar protein concentrations with a novel type of nanowire-based biosensor. The biosensor is based on an electrostatically formed nanowire, which is conceptually different from a conventional silicon nanowire in its confinement potential, charge carrier distribution, surface states, dopant distribution, moveable channel and geometrical structure. This new biosensor requires standard integrated-circuit processing with relaxed fabrication requirements. The biosensor is composed of an accumulation-type, planar transistor surrounded by four gates, a backgate, front gate and two lateral gates, and it operates in the all-around-depletion mode. Consequently, adjustment of the four gates defines the dimensions and location of the conducting channel. It is shown that lithographically shaped channels of 400 nm in width are reduced to effective widths of 25 nm upon lateral-gate biasing. Device operation is demonstrated for protein-specific binding, and it is found that sensitive detection signals are recorded once the channel width is comparable with the dimensions of the protein. The device performance is discussed and analyzed with the help of three-dimensional electrostatic simulations.