In recent years, the development of nanophotonic devices has presented a revolutionary means to manipulate light at nanoscale. Recently, artificial neural networks (ANNs) have displayed powerful ability in the inverse design of nanophotonic devices.
However, there is limited research on the inverse design for modeling and learning the sequence characteristics of a spectrum. In this work, we propose a novel deep learning method based on an improved recurrent neural networks to extract the sequence characteristics of a spectrum and achieve inverse design and spectrum prediction. A key feature of the network is that the memory or feedback loops it comprises allow it to effectively recognize time series data. In the context of nanorods hyperbolic metamaterials, we demonstrated the high consistency between the target spectrum and the predicted spectrum, and the network learned the deep physical relationship concerning the structural parameter changes reflected on the spectrum. Moreover, the proposed model is capable of predicting an unknown spectrum based on a known spectrum with only 0.32% mean relative error. We propose this method as an effective and accurate alternative to the application of ANNs in nanophotonics, paving way for fast and accurate design of desired devices.
We present a monolayer black phosphorus (BP)-based metamaterial structure for tunable anisotropic absorption in the mid-infrared. Based on the critical coupling mechanism of guided resonance, the structure realizes the high absorption efficiency of 9
9.65$%$ for TM polarization, while only 2.61$%$ at the same wavelength for TE polarization due to the intrinsic anisotropy of BP. The absorption characteristics can be flexibly controlled by changing critical coupling conditions, including the electron doping of BP, geometric parameters and incident angles of light. The results show feasibility in designing high-performance BP-based optoelectronic devices with spectral tunability and polarization selectivity.
The use of relatively simple structures to achieve high performance refractive index sensors has always been urgently needed. In this work, we propose a lithography-free sensing platform based on metal-dielectric cavity, the sensitivity of our device
can reach 1456700 nm/RIU for solution and 1596700 nm/RIU for solid material, and the FOM can be up to 1234500 /RIU for solution and 1900800 /RIU for solid material, which both are much higher than most sensing methods. This sensor has excellent sensing performance in both TE and TM light, and suitable for integrated microfluidic channels. Our scheme uses a multi-layers structure with a 10 nm gold film sandwiched between prism and analyte, and shows a great potential for low-cost sensing with high performance.
Black phosphorus (BP), an emerging two-dimensional (2D) material with intriguing optical properties, forms a promising building block in optics and photonics devices. In this work, we propose a simple structure composed of BP sandwiched by polymer an
d dielectric materials with low index contrast, and numerically demonstrate the perfect absorption mechanism via the critical coupling of guided resonances in the mid-infrared. Due to the inherent in-plane anisotropic feature of BP, the proposed structure exhibits highly polarization-dependent absorption characteristics, i.e., the optical absorption of the structure reaches 99.9$%$ for TM polarization and only 3.2$%$ for TE polarization at the same wavelength. Furthermore, the absorption peak and resonance wavelength can be flexibly tuned by adjusting the electron doping of BP, the geometrical parameters of the structure and the incident angles of light. With high efficiency absorption, the remarkable anisotropy, flexible tunability and easy-to-fabricate advantages, the proposed structure shows promising prospects in the design of polarization-selective and tunable high-performance devices in the mid-infrared, such as polarizers, modulators and photodetectors.
