In this paper, dispersion relations (DRs) of photonic crystals (PhCs) are computed by multilayer perceptron (MLP) and extreme learning machine (ELM) artificial neural networks (ANNs). Bi- and tri-dimensional optimized structures presenting distinct DRs and photonic band gaps (PBGs) were selected for case studies. Optical properties of a set of PhCs with similar geometries and different dimensions were calculated by an electromagnetic solver in order to provide input data for ANN training and testing. We demonstrate that simple- and fast-training ANN models are capable of providing accurate DRs' curves in a very short time.

Optical nanoantennas, especially of the dipole type, have been theoretically and experimentally demonstrated by many research groups. Likewise, the plasmonic waveguides and optical circuits have experienced significant advances. In radio frequencies and microwaves a category of antenna known as dielectric resonator antenna (DRA), whose radiant element is a dielectric resonator (DR), has been designed for several applications, including satellite and radar systems. In this letter, we explore the possibilities and advantages to design nano DRAs (NDRAs), i. e., DRAs for nanophotonics applications. Numerical demonstrations showing the fundamental antenna parameters for a circular cylindrical NDRA type have been carried out for the short (S), conventional (C), and long (L) bands of the optical communication spectrum.

Future expansion of computing capabilities relies on a reduction of energy consumption in silicon-based integrated circuits. A promising solution is to replace electrical wires with optical connections, for which a key component is a nanolaser that coherently emits into silicon-based waveguides to route information across a chip, in place of bulky off-chip devices. We report room temperature, sub-${\mathrm{\mu m}}^2$ footprint, quantum-well-in-nanopillar lasers grown directly on silicon and silicon-on-insulator (SOI) substrates that emit within the silicon-transparent wavelength range under optical excitation. The laser wavelength is controlled by changing the InGaAs quantum well thickness and alloy composition, quite independent of lattice mismatch with the InP barrier, a unique property of the 3D core-shell growth mode. We achieve excellent luminescence yield and low continuous wave transparency power due to the well-passivated InGaAs/InP interfaces. These sub-${\mathrm{\mu m}}^2$ footprint long-wavelength lasers could enable optoelectronic integration and photon routing with silicon waveguides on the technologically relevant SOI platform.

The prediction of dispersion relation and photonic band gaps in 2-D photonic crystals using artificial neural networks is demonstrated in this letter. Two case studies are carried out in order to evaluate the advantages of using this approach in conjunction with numerical methods. The results obtained present values that are very close to those obtained by a numerical solver and with short time-processing. We also compare artificial neural networks' outputs and well-known interpolation techniques.

Highly compact III–V compound semiconductor active nanophotonic devices integrated with silicon are important for future low power optical interconnects. One approach toward realizing heterogeneous integration and miniaturization of photonic devices is through nanowires/nanopillars grown directly on silicon substrates. However, to realize their full potential, the integration of nanowires/nanopillars with silicon-based electronics must be made scalable via precise control of nanopillar site and dimensions. Here we demonstrate the first electrical-pumped InGaAs/InP multiquantum-well (MQW) light emitting diodes (LED) using nanopillar array directly grown on a Si substrate with site control, with current conduction directly through the silicon. The growth is via catalyst-free, low-temperature metal organic chemical vapor deposition, which is CMOS compatible. We report excellent optical properties including long minority carrier lifetimes and room-temperature lasing under optical pumping. InGaAs/InP quantum wells are incorporated in the nanopillars in a core–shell growth mode, to obtain silicon transparent emission of ∼1510 nm with high internal quantum efficiency (∼30%). Despite its small footprint, a high output power (4 μW) was measured, and the device could be electrically biased to produce optical gain. CMOS-compatible site-controlled growth and electrically driven long-wavelength emission make the InP nano-LED an ideal component in advanced photonic integrated circuits.