ABSTRACT
A compact, ultra-broadband and high-performance silicon TE-pass polarizer is proposed and demonstrated experimentally. It is based on partially-etched (ridge) waveguide adiabatic S-bends, input/output tapers and side gratings on a silicon-on-insulator (SOI) platform. A compact footprint and weak back reflections are obtained due to the bent waveguide and the tapers, respectively. An extremely high extinction ratio is achieved by scattering the undesired light in the slab section using the side gratings. The 3D FDTD simulations show a TE loss less than 0.3 dB and an extinction ratio greater than 30 dB over a 500 nm wavelength range (1200 nm to 1700 nm). Measured results show a high TM loss (> 35 dB) and a low TE insertion loss (< 1.5 dB), over a 200 nm wavelength range (1450 nm to 1650 nm). The measured TE loss is < 0.6 dB at a communication wavelength of 1550 nm. The footprint of the optimized design is 65 µm × 20 µm.
ABSTRACT
Artificial neural networks have shown effectiveness in the inverse design of nanophotonic structures; however, the numerical accuracy and algorithm efficiency are not analyzed adequately in previous reports. In this Letter, we demonstrate the convolutional neural network as an inverse design tool to achieve high numerical accuracy in plasmonic metasurfaces. A comparison of the convolutional neural networks and the fully connected neural networks show that convolutional neural networks have higher generalization capabilities. We share practical guidelines for optimizing the neural network and analyzed the hierarchy of accuracy in the multi-parameter inverse design of plasmonic metasurfaces. A high inverse design accuracy of $\pm 8\;{\rm nm}$±8nm for the critical geometrical parameters is demonstrated.
